Adoption of Scanning Technology in the Secondary Wood Products Industry

by

Dan W. Cumbo II

Masters Thesis submitted to the Faculty of the Virginia Polytechnic Institute and State University in partial fulfillment of the requirements for the degree of:

Masters of Science

in

Wood Science and Forest Products

Approved:

__________________________ ____________________________ Dr. D. Earl Kline, Co-Chair Dr. Robert L. Smith, Co-Chair

___________________________ Dr. Fred M. Lamb

ii

Adoption of Scanning Technology in the Secondary Wood Products Industry

Dan W. Cumbo II

Graduate Research Assistant

ABSTRACT

A mail survey of wood cabinet, furniture, dimension, and flooring manufacturers

was undertaken across the US to assess the market potential for automated lumber

grading technology in the secondary wood products industry. Two hundred and nine

usable responses were received providing information related to the identification of

adopters of scanning technology, the optimum benefit bundle to increase the marketable

success of automated lumber grading technology, and the best methods of promotion to

this industry.

The results indicate that potential adopters of automated lumber grading

technology will most likely be larger companies both in terms of annual sales and number

of employees. In addition, adopter companies will be more likely to market their

products in international markets. Results show that dimension and flooring

manufacturers perceive this technology as more beneficial to their industries than cabinet

and furniture manufacturers.

Analysis showed that significant differences existed between the four industry

sectors on various machine attributes. In addition, dimension and flooring manufacturers

rated machine attributes higher than cabinet and furniture manufacturers where

significant differences were detected. This indicates that dimension and flooring

iii

manufacturers place more importance on various machine attributes than cabinet and

furniture manufacturers when considering the adoption of automated lumber grading

technology. Optimal benefit bundles were characterized for each industry sector. While

these benefit bundles differed among sectors, the differences where slight. Machine

accuracy, simplicity of operation, technical support, and reduction in labor costs were

important attributes to most sectors.

The best methods of promotion to the secondary wood products industry included

trade shows, peer discussions, and plant visits. In addition, these three methods were

most effective in promotion for the individual industry sectors as well. Advertisements,

meetings and symposiums, scientific journals, and unsolicited sales literature were not

rated as highly effective methods of promotion to the cabinet, furniture, dimension, and

flooring industries.

iv

DEDICATION

This study is dedicated to my wife, Stacy E. Cumbo, who has provided me with

love and support through the most difficult of times. From the time we met she has given

me unconditional love and support without fail. She has always been able to help me see

the bright side of every situation. In addition, I would like to recognize my parents,

Danny W. Cumbo Sr., Ruth S. Hankins, and Sue W. Cumbo. I thank them all for helping

me become the man I am today. Also, many thanks to my family for giving me those

occasional words of encouragement that help in ways that cannot be described. Thank

you all.

v

ACKNOWLEDGEMENTS

I would like to thank the following people for their support of this project:

My most sincere appreciation to Dr. D. Earl Kline and Dr. Robert L. Smith for providing me with the opportunity to work on such a challenging and rewarding project. They have provided immeasurable support, assistance, and guidance throughout the past sixteen months. My sincere appreciation to Dr. Fred M. Lamb for serving on my graduate committee and providing me with answers to many questions. My sincere appreciation to Dr. Geza Ifju, who gave me the opportunity to continue my education. My thanks to Joanne Buckner and Susan Gibson for their friendship, and help with this project. My thanks to my fellow graduate students: Chris Surak, Scott Bowe, David Olah, Daryl Corr, Ren-Jye Shiau, Delton Alderman, Nathan Hosterman, David Carradine, Stephanie Gomon, and Pete Urhig for their friendship and support, and for creating the necessary diversions to work that are helpful in a project of this magnitude.

vi

PREFACE

This thesis consists of four sections. Chapter One provides justification for this

research, defines the objectives of this research, and reviews literature relevant to the

adoption of scanning technology in the secondary wood products industry. The

remainder of the thesis consists of three sections. Chapter Two discusses overall industry

demographic information used to characterize the market as a whole, and characteristics

of potential adopters of automated lumber grading technology. Chapter Three discusses

the benefit bundle that will increase the overall marketable success of automated lumber

grading technology in the secondary wood products industry, and the best methods of

promotion to this industry. Chapter Four consists of a summary of the results and

conclusions, prescribes recommendations for future research, and outlines the limitations

to this study. In addition, some duplication of information between manuscripts, to allow

them to stand alone, was unavoidable. The author apologizes for any inconvenience this

may cause the reader.

vii

TABLE OF CONTENTS abstract ................................................................................................................................ ii dedication ........................................................................................................................... iv Acknowledgements ............................................................................................................. v Preface................................................................................................................................ vi TABLE OF CONTENTS..................................................................................................vii LIST OF TABLES AND FIGURES.................................................................................. ix

Tables ............................................................................................................................. ix Figures............................................................................................................................ ix

Chapter 1 ............................................................................................................................. 1 abstract ................................................................................................................................ 1 INTRODUCTION............................................................................................................... 2 Problem Statement and Justification ................................................................................... 6 Problem Definition.............................................................................................................. 8 Research Objectives ............................................................................................................ 9 Literature Review.............................................................................................................. 10

Secondary Wood Processing......................................................................................... 10 Lumber Grading ........................................................................................................ 10 Furniture .................................................................................................................... 12 Dimension and Wood Components........................................................................... 13 Cabinets..................................................................................................................... 14 Flooring ..................................................................................................................... 15 Recovery and Yield................................................................................................... 16

Development of Optimization Systems......................................................................... 17 Current Scanning Technology....................................................................................... 19 Marketing Issues ........................................................................................................... 20

Industrial Product Introduction ................................................................................. 22 High Technology Products........................................................................................ 25 Innovation Attributes................................................................................................. 27 Product Bundling....................................................................................................... 28

Methodology ..................................................................................................................... 30 Sample Development .................................................................................................... 31 Questionnaire Outline ................................................................................................... 33 Data Analysis ................................................................................................................ 34 Non-response Bias......................................................................................................... 38

Project Summary............................................................................................................... 39 rEFERENCES ................................................................................................................... 41 Chapter 2 ........................................................................................................................... 45 Abstract ............................................................................................................................. 45 Objectives.......................................................................................................................... 50 Methodology ..................................................................................................................... 50 Results ............................................................................................................................... 51

Market Characterization................................................................................................ 52

viii

Differences among Adopters and Non-adopters ........................................................... 58 Conclusions ....................................................................................................................... 68 references .......................................................................................................................... 71 Chapter 3 ........................................................................................................................... 72 abstract .............................................................................................................................. 72 Introduction ....................................................................................................................... 74 Objectives.......................................................................................................................... 78 Methodology ..................................................................................................................... 78 Results ............................................................................................................................... 79

The Benefit Bundle ....................................................................................................... 80 Promotion...................................................................................................................... 95

conclusions ........................................................................................................................ 99 References ....................................................................................................................... 103 chapter 4 .......................................................................................................................... 104

Research Summary..................................................................................................... 104 Implications to the Secondary Wood Products Industry............................................. 108 Markets for Scanning Technology .............................................................................. 109 Areas of Future Research ............................................................................................ 110 Limitations to the Study .............................................................................................. 111

Appendix A: Survey Questionnaire ................................................................................ 112 Appendix b: Survey Cover Letter ................................................................................... 115 appendix c: reminder postcard ........................................................................................ 116 vitaE ................................................................................................................................ 116 vitaE ................................................................................................................................ 117

ix

LIST OF TABLES AND FIGURES

TABLES

Table 1. NHLA grades ...................................................................................................... 11 Table 2: Percentage of sales and employees per industry sector ...................................... 54 Table 3: Percentage of international marketing, company ownership, scanning tech

adoption according to sector ..................................................................................... 55 Table 4: Percentage of man-hour costs of grading lumber, payback period for automated

lumber grading system according to industry group................................................. 57 Table 5: Percentage of companies finding automated lumber grading system beneficial,60 Table 6: Mean man-hour cost of grading lumber according to industry sector ................ 61 Table 7:Relation of adoption of scanning ......................................................................... 65 Table 8: Comparison of past adopters vs. non-adopters on system attributes .................. 66 Table 9: Mean ratings of methods of leraning about new technologies, adopters ............ 68 Table 10: Mean ratings of machine attributes according to industry sector ..................... 86 Table 11: Optimum benefit bundle according to .............................................................. 88 Table 12: Mean ratings of factors in lumber grading according to industry sector .......... 89 Table 13: Percentage of adoption factors, location of lumber grading, and expected

machine ..................................................................................................................... 92 Table 14: Mean ratings of methods of learning about new technology according to sector

................................................................................................................................... 97 Table 15: Important methods of learning about new technology...................................... 98

FIGURES

Figure 1: Overview of AGS ................................................................................................ 9 Figure 2: Mean lumber production volumbe according to industry group ....................... 53 Figure 3: Mean man-hour cost of lumber grading according to industry group ............... 56

1

CHAPTER 1

ABSTRACT

The forest products industry is becoming a high-tech industry. As technology advances, and

technologies from other industries become available, many companies are searching for ways to automate

their operations, increase efficiency, and improve quality. Scanning technology, first introduced to the

softwood lumber industry in the 1970s, has more recently been utilized in hardwood lumber production and

secondary manufacturing of wood products. Scanning systems are now being used in automated chop saw

operations, quality control systems, and other wood optimizing systems. Current research and product

development projects are focusing in the area of defect detection scanning systems. These systems are

designed to distinguish clear wood from defects such as knots, wane, and decay. Information generated can

be used to saw lumber for optimum yield, better match lumber to specific products, or assign grades to

lumber. This study investigated the differences in factors affecting the adoption of technology, among four

segments of the secondary wood processing industry. The four sectors investigated include furniture

manufacturers, cabinet manufacturers, wood dimension manufacturers, and flooring manufacturers. In

addition, this research was designed to characterize adopters from non-adopters, and predict the best

methods to promote scanning technology to each industry segment

2

INTRODUCTION The secondary wood products industry is comprised of a variety of manufacturers

utilizing a wide range of species and grades of wood. Greater than 90% of all hardwood

lumber produced in the US is consumed by eight secondary wood processing markets.

These markets include pallets, furniture, dimension parts, exports, millwork, flooring,

cabinets, and railway ties. Hardwood lumber consumption for these markets has been

estimated at over 13 billion board feet in 1997 (Hansen and West 1998). Consumption

figures for 1997 show an increase of over two billion board feet since 1991 (Dempsey

and Luppold 1992). In the eight markets previously mentioned, only the pallet industry

failed to show an increase in hardwood lumber use in this decade. With the strong

economic climate in the US expected to persist, continued increases can be expected.

While the wood products industry has not been traditionally known for its

utilization of high technology equipment, many within the industry are seeking methods

to automate their processes to increase lumber yield and product value. In addition, many

companies are searching for methods to regulate the precision of their processes, improve

their quality control operations, and facilitate the tallying and sorting of materials.

Patterson et al. (1997) noted that 36% of furniture manufacturers and 55% of cabinet

companies had some type of computer numeric control (CNC) machinery in place. In

addition, greater than 39% of the furniture companies and 13% of the cabinet

manufacturers indicated that they planned to purchase a CNC machine within 12 to 24

months. The most common types of CNC machinery observed were routers, carving

3

machines, double-end tenoners, sanders, moulders, shapers, boring machines, and

grinders.

Scanning technology, which was first introduced into the softwood lumber

industry some twenty years ago as a component of yield optimization systems, has more

recently been utilized in the secondary wood products segment as well. Currently, a great

deal of research persists in defect detection and specialized interpretation and utilization

of data generated by defect detection systems (Pham and Alcock 1998).

Laser scanning is used to locate such aspects as wane, voids, and board shape.

Color and black and white camera scanning technology can be implemented in color

sorting and matching in processes such as cabinet door panel production, as well as in the

identification of mineral streaks, surface knots, and pith. X-ray scanning technology can

be used to detect internal defects such as internal knots, internal voids, and honeycomb

(Kline et al. 1998). Other scanning technologies exist which detect and utilize

differences in wood cell structure to identify defects in lumber.

Lumber scanning technology for grading and processing hardwood lumber is now

in the development stages. It is expected to become more commonplace in the industry

as the proficiency of the technology advances and is proven to the industry. Current

justification for purchasing a scanning system is the reduction of labor. However, the

value of information that can be received from scanners may be of more importance.

Kline et al. (1998) state that “The primary cost savings from an automated lumber

grading system will be realized by producing a more uniform and consistently graded

product and by producing a higher value product through optimum lumber

4

remanufacturing.” The ability to optimize yield and value are perhaps the most

important attributes of lumber scanning systems, however, these systems are limited by

the information they receive (Mullin 1998). This implies that the algorithm generated by

an automated grading system is only as reliable as the information received from the

scanning heads.

It is obvious that the adoption of lumber scanning technology by the secondary

wood products industry depends largely on the performance of these systems in actual

industrial applications. Rosenberg et al. (1990) noted that technology adoption in the

forest products industry is influenced by the role of innovations imported from other

industries, the effect of raw material shortages, the importance of economic factors in

adoption of innovations, and problems associated with the heterogeneity of wood raw

material and finished products.

Technology is often developed at the expense of one industry, while other

industries later enjoy the benefits of the technology. For example, x-ray technology has

been used in the medical field for a number of years, and in the case of automated lumber

grading systems, it is now diffusing into the forest products industry. In addition, wood

products manufacturers will focus more on improving processing operations in periods of

economic recession and at times when raw material quality is not consistent. These

scenarios lead to greater emphasis on increasing product value and part yield, and result

in more innovation within the industry.

From the marketing perspective, the adoption of industrial high technology

products can be evaluated on their perceived attributes. Mittal et al. (1998) state that

5

negative performance on an attribute has a greater impact on overall satisfaction than

positive performance on the same attribute. In addition, overall customer satisfaction

shows diminishing sensitivity to attribute-level performance. By installing systems that

possess attributes desirable to specific manufacturing operations, overall satisfaction level

will increase.

Lehmann and O’Shaughnessy (1974) define industrial products as Type I, II, III,

or IV products. A Type I product is routinely ordered and used. Type I products are

expected to do the job and there is no question as to how to use it. “For Type II products,

the buyer is confident that the product can do the job. However, problems are likely,

because personnel must be taught how to use the product” (Lehmann and

O’Shaughnessy 1974). The buyer will favor the supplier whose offering is perceived as

likely to minimize the time and difficulty associated with learning the operation of the

product. “With Type III products, there is doubt as to whether the product will perform

satisfactorily in the application for which it is being considered. It is argued that the

buyer will favor the supplier who can offer appropriate technical service, provide a free

trial period, and who appears flexible enough to adjust to the demands of the buyer’s

company” (Lehmann and O’Shaughnessy 1974). Type IV products may result in

problems within the company. With Type IV products, it is likely that there will be

difficulty in reaching agreement among those affected by adoption of the product. Defect

detection, lumber scanning systems display characteristics of both Type II and Type III

products. Manufacturers of products with these characteristics must provide systems that

are simple to operate. In addition, the company must be flexible relative to buyer

6

demands. This research is designed to identify the specific attributes that will increase

the marketable success of automated lumber grading technology. There has been no

research conducted, which addresses scanning technology adoption factors in the

cabinets, furniture, dimension, and flooring industries.

PROBLEM STATEMENT AND JUSTIFICATION Secondary wood products manufacturers are facing the problem of continually

increasing production efficiency and part yield to remain competitive. As resources

become scarce, competition increases due to expanding global competition, and

alternative materials continue to target markets historically dominated by wood,

manufacturers of wood products must attempt to improve their production processes.

Process automation, increased efficiency, increased yield, and improved information

exchange will become issues of growing importance throughout the industry.

Christianson (1998) showed levels of automation and machine purchases among wood

products manufacturers to be increasing into the next millenium. The top industry

concern in this study was recruiting and retaining skilled workers. It was also noted that

more large companies classified their shops as technologically above average while

smaller companies tended to view themselves as below average. In addition, among most

companies, the need to up-grade technology was not only caused by competitive and

market pressures, but also by reliance on a diminishing pool of skilled workers who

require salaries at levels many companies cannot offer.

7

The secondary wood manufacturing industry is made up of a variety of producers

manufacturing a heterogeneous mix of products. To remain competitive in a more

sophisticated global marketplace, secondary manufacturers must become increasingly

technologically advanced thereby increasing efficiency and optimizing product yield.

Yield is perhaps the most important aspect of the manufacturing process for most forest

products firms. Huber et al. (1985) state that analyses of furniture plant operations have

shown that improvements in yield are key to the success of the rough mill. The affect on

yield is a significant concern when forest products firms consider changes in the

manufacturing process. Furthermore, firms must be aware of potential marketing

implications associated with technology adoption. Innovation and technology adoption

may result in changes in marketing strategy.

“Most policy makers believe adoption of advanced technologies is fundamental to

wood manufacturers’ competitiveness in global markets” (Hoff et al. 1997). In

manufacturing, competitiveness is heavily influenced by productivity, product quality,

and the introduction of new products. Some progressive furniture manufacturers have

dealt with the threat of foreign competition by automating their processes and thereby

becoming more efficient. In a labor-intensive industry, automation typically results in

loss of jobs. However, subsequent increases in productivity and efficiency have allowed

some companies to expand operations creating other jobs that offset lost jobs. In

addition, Hoff et al. (1997) suggest that firms who possess advanced manufacturing

technologies tend to be more proficient in responding to changing customer demands.

8

PROBLEM DEFINITION

The overall goal of this study is to understand the factors that influence the

adoption of emerging technologies in the secondary wood products sector. The four

segments of the secondary wood products industry selected for this project include

furniture, cabinets, dimension, and flooring manufacturers. They were chosen based on

the relative importance of lumber grading in their processing operations. These segments

represent the most value-added products in the secondary wood products industry. Other

products in this industry including railway ties, millwork, and pallets do not grade

sufficient amounts of lumber to be relevant to this study. The emerging technology,

considered in this study, is automated lumber grading technology (AGS) (Figure 1).

Research is currently being conducted in the area of automated lumber grading

technology. (Kline et al. 1998, Pham & Alcock 1998, and Mullin 1998). It is important

for manufacturers of automated lumber grading systems to understand the market

potential, “benefit bundle,” and methods of promotion to bring this technology to market.

The term “benefit bundle” is defined in this study as the system capabilities and product

attributes offered by automated lumber grading system manufacturers and/or required by

secondary wood manufacturers. It is thought that a lack of productive communication

and understanding exists between manufacturers of scanning technology and secondary

wood processing companies.

9

MaterialsHandlingControl

DetectLumber

ColorCamera

1

ColorCamera

2

RangingCamera

1

RangingCamera

2

X-RaySensor

Image ProcessingComputer

ControllableStop

ControlComputer

X-RaySource

Figure 1: Overview of AGS

RESEARCH OBJECTIVES

This study is designed to characterize the potential market for an automated

lumber grading system in the four specified sectors of the secondary wood products

industry. In determining market potential, other characteristics of the market are

investigated to accurately estimate potential adoption of the product. The specific

research objectives are:

1. to identify factors which characterize firms as potential adopters or non-adopters of automated lumber grading technology;

2. to characterize the “benefit bundle” that will increase the marketable success

of an automated lumber grading system by secondary wood processors;

10

3. to identify the best methods of promotion to secondary manufacturers of wood products.

LITERATURE REVIEW

SECONDARY WOOD PROCESSING

Secondary processors of hardwood lumber add value through processing beyond

the rough green lumber stage. Lumber serves as the input raw material, which is further

processed into a new value-added product. Products such as wood cabinets, furniture,

dimension, flooring, millwork, and railway ties represent secondary wood products. In

1997, roughly 52% of all hardwood lumber produced in the US was consumed by four

sectors of the secondary wood products industry. These sectors include furniture,

dimension, cabinets, and flooring. Hardwood lumber consumption in 1997 was estimated

at over 7.7 billion board feet among these four sectors (Hansen and West 1998).

Lumber Grading

Hardwood lumber is sold according to grade and volume. The universal grading

system, used in the US and abroad, is based on a set of rules developed by the National

Hardwood Lumber Association (NHLA) (Table 1). The rules were written with the end

user in mind (NHLA, 1999).

The highest grade of lumber under NHLA rules is First and Seconds (FAS). FAS

is also the most expensive grade of lumber. Below the FAS grade are the FAS-1 Face

and Selects grades. These are the only grades that are not based on the poorest face of

the board. The FAS-1 Face and Selects grades indicate that only one face of the board

can meet FAS specifications, while the other face must be graded 1C. These two grades

11

are often grouped together and sold as Face & Better, or Selects & Better lumber. The

Selects grade differs from the FAS-1 Face in that Selects are subject to lower board size

constraints (Table 1). Following FAS-1 Face and Selects in value is the No. 1 Common

(1C) grade. No. 1 Common lumber is the average grade of lumber sold. No. 2 Common

(2C) is the next most valuable grade, and No. 3 (3AC) is the lowest grade for hardwood

lumber.

Table 1. NHLA grades

Grade

Minimum board length

Minimum

board width

Minimum cutting

size

Minimum

area of clear

cuttings FAS 8’ 6” 4”x 5’ or 3”x 7’ 83%

Selects

6’

4”

4”x5’ or 3”x7’ FAS 4”x2’ or 3”x3’ 1C

83% FAS 67% 1C

1C 4’ 3” 4”x 2’ or 3”x 3’ 67% 2C 4’ 3” 3”x 2’ 50%

3AC 4’ 3” 3”x 2’ 33%

The most subjective aspect of lumber grading is determining what is and what is

not a defect; i.e., knot, decay, wane, etc., so that clear cutting areas can be established.

Lumber is graded from the poorest side of the board in terms of defects, with the

exception of Selects or FAS-1 Face. Misinterpretation of defects is perhaps the most

influential on grading accuracy.

Lumber is graded at the sawmill according to NHLA grading rules. In most

secondary wood processing operations lumber is delivered from the sawmill at which

time it is again graded by a certified grader employed by the receiving company. The

processing operation and individual company policy determines the point at which

12

lumber is graded. Each company, depending on the process, will have a target acceptable

grade mix. Any material that does not meet the acceptable standards is typically sorted

out for return to the sawmill. At the grading station, a certified grader inspects each

board keeping a grade tally, which is then used to verify invoices from the sawmill.

Inaccurate grading has varying effects on different manufacturing operations,

depending on the process and the product. Therefore, it is important to understand the

different sectors of the secondary wood processing industry to fully understand the

effects. Different sectors within the industry consume varying quantities, grades, and

species of wood.

Furniture

The wood furniture industry showed a 50% increase in hardwood lumber

consumption between 1991 and 1997 (Hansen and West 1998). Important species to the

furniture industry include red oak, yellow poplar, black cherry, hard maple, black walnut,

and white oak. According to Hansen et al. (1995) red oak accounts for approximately

33% of all wood furniture, while yellow poplar makes up roughly 16% of the species

mix. Some softwoods such as white pine, southern yellow pine, and ponderosa pine are

also utilized by the furniture industry. Eastern white pine accounted for 53% of the

softwood lumber used, while ponderosa pine and southern yellow pine were estimated at

18% and 17%, respectively. Furniture manufacturing uses a wide range of lumber grades

as well. In 1993 lumber graded as No. 1 Common made up roughly 60% of the grade

mix for furniture manufacturers. No. 2 Common accounted for approximately 26% and

nearly 12% was designated as FAS and Selects (Hansen et al. 1995).

13

Dimension and Wood Components

The dimension and wood components industry showed an 85% increase in

hardwood lumber consumption between 1991 and 1997 (Hansen and West 1998).

Dimension and wood components are defined as intermediary, value-added products,

which are used in other secondary wood processing operations such as millwork,

furniture, and cabinet manufacturing. “Wood components, such as dimension parts, are

dried and processed to a point where the maximum waste is left at the mill and the

maximum utility is delivered to the customer.” (Vlosky 1996). In some cases furniture

manufacturers own and operate their own dimension mills, and often dimension mills are

affiliated with sawmills, but most are independent companies. According to Vlosky,

(1996) dimension manufacturers are often the focus of state economic development

initiatives in an attempt to add value to hardwood resources.

Dimension and component operations utilize a wide range of lumber grades and

species. Vlosky (1996) showed red oak to be the predominant species at roughly 39% of

the species mix. Following red oak were yellow poplar, white oak, maple, and cherry at

16%, 15%, 10.5%, and 5%, respectively. The study also noted that geographically, the

North Central US dominated in red oak, while the Northeast was the largest user of

cherry, and the South used mostly poplar, white oak, and maple. Hass and Smith (1997)

showed hardwood components made from red oak accounted for 35% of the value of

components sold in 1994. Following red oak was hard maple at 16%, yellow poplar at

12%, white oak at 10%, cherry at 6%, ash at 5%, and hickory at 3%. It was also noted

14

that the two largest domestic and international end users of dimension and component

parts were wood furniture, and moulding/millwork manufacturers.

Cabinets

From 1991 to 1997 hardwood lumber consumption in the cabinet industry

increased by 33% (Hansen and West 1998). According to Punches et al. (1995) the US

cabinet industry consists primarily of a relatively small number of large companies. In a

study by Adams (1999) it was noted that cabinet sales for use in rooms other than

kitchens and bathrooms has steadily increased throughout the 1990s. Eighteen of 36

cabinet companies surveyed responded that they sell other-room cabinets. In addition, it

was noted that the cabinet industry has experienced 33 consecutive months of growth in

sales since 1996, and both stock and custom manufacturers have averaged double-digit

growth in 1998.

Oak is credited with approximately 80% of solid wood use. Cherry, maple, and

birch follow with an average use of roughly 4% to 8%. Pine is estimated at less than 2%

of total solid wood use. Doud et al. (1992) noted that the most common lumber

purchasing strategy for cabinet manufacturers in Pennsylvania was to buy lumber graded

as FAS only. However, this strategy may differ according to geographic region and raw

material availability.

Other grade mixes included FAS and Selects, FAS and No. 1 Common, No. 1

Common and Selects, or No. 1 Common. These lumber purchasing strategies were less

common than the FAS only strategy in the study by (Doud et al. 1992). In addition, short

length lumber, 4 to 8 feet in length, was shown to dominate due to lower prices. Non-

15

conformity to National Hardwood Lumber Associations grading rules was cited as the

reason for lower prices of short length lumber (Doud et al. 1992).

Smaller companies in the cabinet manufacturing industry tended to use a mixture

of solid wood, wood-veneered, and non-veneered panels while larger companies used

more non-veneered panels. This can be explained by the increased emphasis on stock

cabinets in larger companies compared to smaller companies who place more emphasis

on custom cabinets (Punches et al. 1995).

Flooring

From 1991 to 1997 hardwood lumber consumption in the flooring industry

increased by 120% (Hansen and West 1998). Recently wall-to-wall carpet and PVC

flooring have begun to lose their market share to wood flooring as a result of increasing

health and environmental concerns among consumers (Weinig 1999). Weinig states,

“Reservations about chemicals and fear of house dust by many allergy sufferers has led

to a decrease in carpet sales versus wood flooring.”

Wood flooring utilizes a variety of species and grades. The National Wood

Flooring Association lists oak, ash, hard maple, beech, birch, and pecan as the most

significant species. However, other species are used such as locust and some pine species

(NWFA, 1999). In addition, there are several types of wood flooring to choose from.

These include unfinished, pre-finished, solid wood, wood laminate, and acrylic

impregnated wood flooring. Solid wood flooring is produced in strip form, plank form,

or as parquet flooring. Strip flooring is manufactured to ½” or ¾” thickness and widths

of 1 ½”, 2”, and 2 ¼”. Plank flooring is manufactured to greater widths than strip

16

flooring, 3” to 8”, and parquet flooring is made up of geometric patterns, which are

composed of individual wood slats. Parquet flooring is typically manufactured from

short pieces 6 to 12 inches in length. Unlike the wide range of grades utilized by other

secondary manufacturers, flooring manufacturers compete for a grade mix comprised

mostly of lower grade material (Hansen and West 1998). In general, flooring

manufacturers utilize primarily No. 1 Common and lower grades of wood.

Recovery and Yield

Perhaps the most important issue concerning manufacturers of forest products

today is recovery or yield obtained from raw materials, logs for primary wood processors,

and lumber in secondary wood processing. Secondary manufacturers of wood products

are constantly evaluating the yield percentages achieved in the mill. Furthermore, most

mill managers are striving to at least maintain a certain level, if not improve, yield. Yield

can be considered a measure of efficiency, a reflection of productivity, and a potential

area of savings for both the company and the customer. It has been noted that improving

yield is the key to rough mill success (Huber et al. 1985). There are a number of

techniques for measuring yield, but ultimately every mill has a set percentage they

attempt to maintain. For example, a furniture manufacturer may be satisfied with roughly

60% lumber utilization in the production process, while a dimension mill may expect a

higher yield percentage. It should be noted that with a yield of 60%, 40% of the raw

material is not being utilized in the form of a final product. This presents an area for

significant potential savings. Many operations are looking to optimizing scanning

systems to obtain the most usable material possible. Huber et al. (1985) suggest that the

17

ultimate solution for maintaining productivity and yield in the rough mill is the virtual

elimination of human error through use of computer-aided manufacturing methods and

technology such as computer vision.

DEVELOPMENT OF OPTIMIZATION SYSTEMS

During the lull in the market for forest products in the late 1970s and early 1980s,

many sawmill operations began to feel the need to embrace automation, computer

control, and overall process management more vigorously. Softwood lumber

manufacturers realized that by precisely controlling processing operations using

automated control systems they could increase lumber recovery and more easily adjust to

changes in demand. In the early 1980s some Canadian firms began using light emitting

scanners with cameras and integrated circuits to measure log dimensions prior to

bucking. In addition, scanning systems were introduced as components of chipping-

canting systems for pre-setting chipping heads (Saunderson 1982).

Scanning technology gained significance in the forest products industry as

optimization systems used for primary processing of softwood timber. Optimization

technology did not gain significance in the hardwood lumber industry until the late

1980s. The first edger-optimizer designed specifically for hardwood production was

installed in a sawmill in Pennsylvania in 1988 by INOVEC. Edger/trimmer optimizing

systems can be separated into three basic components; the scanner, the decision-making

software, and the mechanical apparatus that carries out the solution. Early edger-

optimizers claimed 92% recovery compared to the 70% to 85% that was common with

manual systems. The first edger-optimizer in the US was installed at Cottage Grove, OR,

18

in a small log softwood sawmill owned by Weyerhaeuser Co. in 1976. The mill was

touted as Weyerhaeuser’s most automated small log mill (White 1977). The system was

developed and built in Sweden, and the purchase cost was listed at $285,000. Boards

were scanned by a photodiode array, and edged according to data stored in computer

memory.

Another significant contributor to the increase in use of scanning systems, and

automation in general, was the increasing trend to manufacture lumber from smaller logs

in the 1970s. As environmental regulations changed and smaller second growth trees

were harvested, log processing operations were forced to adapt. Williston (1979)

suggested that the Best Opening Face Method (BOF) of log breakdown, developed by the

USDA Forest Products Laboratory at Madison, Wisconsin, resulted in further advances in

scanning technology, computing, and positioning devices. Today, the most widely used

log scanning systems measure only external features employing log profile sensing

techniques (Kline 1998).

In 1977 the Forest Products Lab in Madison developed a defect scanning system

known as the Defectoscope, which used ultrasound to detect irregularities in wood grain

(Anonymous 1977). Sound waves were passed through the board detecting defects such

as knots and decay. A computer directed a line plotter to reproduce the image of the

board indicating where defects were located, then marked where cuts should be made for

maximum yield. Potential yield increases using the Defectoscpoe were calculated at

15%. Based on this data, it was noted that a $100,000 investment could be paid off

19

within six months from the increased yield alone in a medium sized mill (Anonymous

1977).

Hardwood lumber grading differs from softwood lumber grading in that

hardwood lumber is graded based on aesthetic quality rather than strength limiting

defects, therefore, hardwood lumber production and grading focus more on product value

than volume. In the past hardwood lumber scanning systems have been based on a single

sensing modality; e.g., black and white or color camera, sensing only gross wood features

(Kline 1998). “In secondary lumber processing, a completely automatic lumber

inspection system must not only accurately detect the size and shape of lumber, the

system must also accurately find the location and type of defect with sufficient precision

and resolution” (Kline 1998).

CURRENT SCANNING TECHNOLOGY

Today, lumber-scanning technology is much more sophisticated and it is currently

in a state of continual development. Much of the current research in lumber scanning is

focused in the area of defect detection. Pham & Alcock (1998) suggest that the speed at

which lumber is graded and the level of stress involved have contributed to the desire to

remove human error from the grading process through automation. Modern secondary

wood products production speeds require lumber grading at higher speeds, however,

accuracy must be maintained. Human error and fatigue associated with manual lumber

grading may result in decreased grade accuracy and precision.

Today, most wood scanning technology utilizes the machine vision system

(MVS) framework. Conners et al. (1999) note that machine vision systems contain the

20

following three features: (1) a combination of electrical hardware and computer software

sensing and measuring spatially varying characteristics of a specimen, (2) sensing

produces an image that can be processed as a 2-dimensional or 3-dimensional array of

measurements, and (3) computer hardware/software automatically interprets those

measurements as an image in order to infer important properties associated with the

specimen. These activities are designed to function with little or no human intervention.

Some manufacturers are now offering systems that promise reduced trim loss,

reduced grader fatigue, and higher percentages of “on grade” packages. Grade scanning

optimizers for chop saw lines are being touted as defect detection systems that have the

ability to optimize based on grade or value. Advertisements claim scanning speeds of up

to 700 lineal feet per minute. In addition, manufacturers are offering systems equipped

with a variety of components in a variety of combinations, while promising new features

as technology advances. The limitation of these systems, as they exist now, is the

decrease in accuracy and precision that results when scanning green rough-surfaced

wood. Higher moisture content lumber causes problems, especially in the color and x-ray

images, due to color and density variation caused by wet pockets in the wood. In

addition, machine and dirt marks on lumber are often falsely recognized as defects by the

scanning cameras (Kline et al. 1998).

MARKETING ISSUES

It is important to note that technology for technology’s sake should not be the

motivation for product development. A technology product, as with any new product,

must supply a need. Furthermore, it must be presented to the market in a timely fashion,

21

and in a manner such that maximum market share can be obtained. “The most serious

mistakes companies make are usually the result of poor planning. By the time a company

finds itself going after the wrong market, with the wrong product, at the wrong time, it is

usually in such a tight box that only Houdini could get it out” (Davidow 1986). Often,

errors in product planning are not apparent until after design and production of the

product. By recognizing at this stage that the wrong market has been targeted significant

loss of resources can occur, both monetarily and in terms of time.

During the early stages of microcomputer development, Intel a well-known

company in the electronics market, developed the successor to the 4004, the world’s first

microcomputer. The company knew the 4-bit 4004 model was too slow, and that an 8-bit

system could be built at no additional cost. In addition, the new model would make use

of several Intel proprietary technologies, which would be difficult for competitors to

duplicate. Customers showed enthusiasm for the new system and the 8048 was born.

However, in spite of all of the euphoria, the 8048 was not successful. Fairchild, a fading

company from which no one expected competition, developed the 3870. Unknown to

anyone, Fairchild had hired a number of talented people with backgrounds in consumer

electronics. They had conducted research on the high volume accounts such as

automobile and telephone manufacturers, and knew what they wanted. Intel had planned

the 8048 for the traditional Intel customer base, which manufactured lower volumes of

more sophisticated systems. The 8048 had many features that were of little value to the

larger consumers, and as a consequence lacked features that were important to them

(Davidow 1986).

22

Intel had developed a technology that was useful only to a small market. It did

produce the intended results and it was an improvement on an existing technology.

However, it was poorly planned and therefore manufactured for the wrong market.

Manufacturers of scanning technology aimed at the forest products industry must

understand the need and potential of the target market to effectively design and promote

the product, and gain market share. “The commercial viability of a new product rests in

the hands of its potential customers; and therefore a solid understanding of the

marketplace together with an effective market launch effort is vital to new product

success” (Cooper 1979).

Industrial Product Introduction

Industrial marketing is defined as the marketing of products and services to

industrial and institutional customers for use in the production of goods and services

(Choffray and Lilien 1980). Marketing to industrial customers differs from marketing to

household consumers in that success in the industrial market depends, to a larger degree,

on other functional areas of the firm. Engineering, research and development,

manufacturing, and technical service all play important roles in the industrial planning

process. Therefore, planning in the industrial market involves greater functional

interdependence and an overall closer relationship to the corporate strategy (Hutt and

Speh 1981). The need for greater functional interdependence when planning industrial

products is based on the complexity often associated with industrial products as well as

increased costs associated with developing a complex or sophisticated product.

23

Deviations from corporate strategy may result in higher development, production, and

marketing costs.

Cooper (1994) states that new product innovation has primarily been a hit-and-

miss affair. However, eight factors driving the success of new products are given. The

factors are as follows:

1. A unique superior product that delivers unique benefits and value to the customer;

2. a strong market orientation, a market driven and customer focused new

product process;

3. up-front research and planning before product design;

4. sharp and early product definition before product development;

5. a cross-functional team approach to new product development;

6. sharp focus and evaluation of time and resource requirements;

7. quality of execution of the innovation process; and

8. a multi-stage-and-gate “Game Plan.”

Summarizing Cooper’s eight factors influencing new product success, the firm

must offer a product for which there is little or no substitute. The new product must be

perceived as superior to similar products and unique in the benefits offered. The firm

must also be market-oriented as opposed to production-oriented. A market-oriented

company has a better understanding of its customer’s needs. Market research must come

before product design to plan the appropriate product for the corresponding market. In

addition, the new product should be clearly defined early in development to help guide

the process. The firm should make efforts to include all areas of the company in the

24

design of new products. This approach allows for insight from people with different

backgrounds in the company. Firms must attempt to accurately estimate resource

requirements in the planning process so development can be kept within budget

constraints. The innovation process must be executed in a timely and deliberate manner.

The product must ultimately be accepted by the customer to be successful. Finally, it is

beneficial to segment the entire planning, developing, and marketing process to more

easily react to unexpected events, and more closely track the process.

The failure of new industrial products has been largely attributed to failure to

adequately manage the product development process. More (1986) suggests that the real

failure has been in understanding and managing strategic relationships with potential

adopting organizations. In addition, structural and process characteristics in both the

developing and adopting organization are strongly linked. Managers in adopting

organizations face similar decision processes as those in the developing organization.

For example, developers must attempt to decide between possible product configurations,

while adopters attempt to choose the most beneficial product among several product

offerings.

Cooper (1979) lists three barriers to the success of new products. These are:

1. Having a high priced product relative to competition with no economic advantage to the customer;

2. being in a dynamic market with many new product introductions; and

3. being in a competitive market, where customers are already well satisfied.

In the case of automated lumber scanning systems, this is a sophisticated industrial

product, which is expected to demand a relatively high price. However, competition is

25

scarce as automated lumber scanning technologies are now emerging. This is a new

product, and as such targeting a new market, therefore market dynamics are developing.

However, as scanning technology is increasingly accepted by the forest products industry

more firms will enter the market.

High Technology Products

Technology, broadly defined, is “know-how,” more specifically it is the

information needed to produce and/or sell a product or service. In business, experience

and information, i.e., know-how are valuable assets. In marketing products, know-how

relative to customer needs is important as well. In this context, know-how may result in a

competitive advantage for the company. Three components of know-how can be

identified: product technology (ideas that are embodied in a product), process technology

(ideas involved in manufacturing a product), and management technology (the set of

procedures associated with business administration and selling a product) (Capon &

Glazer 1987).

Marketers of high technology products must plan the development of products not

devices. Products are designed to fulfill customer needs, while devices are of little or no

use to customers. By virtue of the “high-tech” product classification, these products are

serving new functions in new markets. In a new market firms may not have the luxury of

learning from their competition. There are no lessons to be learned from competitors

who do not exist; therefore, comprehensive planning becomes vital. In high-tech firms,

however, planning often becomes a sidebar to new product development and the

customer is often the last consideration. High-tech companies are reliant upon strong

26

research and development departments to maintain the competitive advantage associated

with being high-tech. The result of over-emphasizing research and development is

typically the development of a product for which there exists little or no market. Often,

an engineer in the laboratory will think of something new. The new idea causes

excitement within the company and a new project is begun. As the program matures,

marketing is expected to find customers who need the new device (Davidow 1986). This

method of product development is the opposite of the marketing-oriented approach where

a need and a market are identified first. The potential of the market is then evaluated

followed by design, promotion, production, and ultimately sales.

Designers must consider the manner in which the product will be used as well as

product durability when developing new high-tech products. Products that are “user-

friendly” have been noted to be more successful than those that are not. In most

situations, a high performance product that is difficult to maintain or operate will see

success only under specific conditions. Furthermore, it is important for managers to

conduct the timing and staging of product design to allow for the intensity of effort

associated with breakthroughs followed by recovery time. In a high-tech market that is

driven by technological advances managers may feel pressure to maintain a sustainable

competitive advantage; however, this may be an unrealistic goal. Perhaps the emphasis

should be placed on maintaining the capabilities to design new product strategies on an

on going basis. For example, the ability to rapidly prototype could facilitate and hasten

the design of new product strategies.

27

Rhyne (1994) suggests that companies wishing to remain competitive in high-tech

markets must:

1. Focus on developing teams that will function together over time;

2. recognize the need for continuous improvement over time despite temporary set backs; and

3. recognize that experience in the industry remains critical.

In high-tech markets it is important that those responsible for product development be

part of a productive working team. Any activity that hinders team focus can be

detrimental to the task at hand. High-tech firms must also be committed to constant

product improvement. As technology advances, manufacturers of forest products will

likely expect automated grading systems to be capable of processing more information

quicker and with less human interaction. In addition, it is important for high-tech firms

to understand that experience in the industry is vital. In time manufacturers of automated

lumber grading systems will come to better understand the forest products industry they

serve. This knowledge may lead to better product development and new machine

features in the future.

Innovation Attributes

Ostlund (1974) suggests that perceived innovation attributes are strong predictors

of new product adoption. The study listed six perceived innovation attributes as

important to new product adoption; i.e., predictors of innovativeness. The attributes are:

1. relative advantage (degree to which an innovation is perceived superior);

2. compatibility (degree to which an innovation is perceived consistent with existing values, habits, and past experiences);

28

3. complexity (degree to which an innovation is perceived too difficult to understand and use);

4. trialability (degree to which an innovation is perceived available on a trial

basis without a large commitment);

5. observability (degree to which the results of an innovation will be apparent and possible to communicate to others); and

6. perceived risk (degree to which risks are perceived to be associated with the

innovation). The significance of perceived innovation attributes in new product adoption reinforces

the importance of planning in high-tech product development. It is important for

managers to understand the ranking of the potential new product on the attributes related

to the product. In addition, specific product attributes play a significant role in the

adoption of high-tech products. In the example of automated lumber grading systems,

producers of different secondary wood products may place varying degrees of importance

on different attributes.

Product Bundling

Often, various product or service attributes are bundled and offered together as a

single product. “Bundling is selling separable products or services to buyers as a

package, or bundle” (Porter 1980). For example, IBM has bundled computer hardware,

software, and technical support for many years. Bundling implies that all buyers are

supplied the same package regardless of differences in need. Buyers often differ in the

intensity of their need for different products and services; therefore, buyer receptiveness

to bundling may vary.

29

Porter (1980) suggests several ways in which bundling can create a competitive

advantage. First, a firm may lower costs by providing a single package rather than a

customer specific mix. Firms can better share activities in the value chain that provides

parts of the bundle. This can result in economies of scale for the firm. Second, bundling

may allow a firm to differentiate itself from other firms selling only parts of the bundle.

In this context bundling can prevent the need to compete with other companies in the

areas of their greatest strength. In addition, compatibility among items in the bundle can

be facilitated if a single firm provides the entire package of goods or services required to

satisfy the customer. In the case of automated lumber grading systems, the compatibility

issue can be important if the market dictates that the system must perform tasks other

than grading. Third, bundling may provide firms with enhanced opportunities for price

discrimination. In cases where buyers have different sensitivities to various parts of the

bundle, individual parts can be offered at prices that sum to more than the bundle price.

Bundling involves risks as well. It makes the assumption that a significant

number of buyers are willing to purchase the bundle; therefore, the diversity of buyer’s

needs is not considered. Lack of attention to specific buyer’s needs may result in

vulnerability to firms that take a more focused approach. Vulnerability may arise if

specialists focus on parts of the bundle, thereby achieving lower costs or differentiation in

producing them (Porter 1980). An automated lumber grading system may be offered as

a bundled product consisting of hardware, up-gradable software, and technical support. It

could be bundled as a product consisting of those attributes in addition to tallying,

sorting, or cut-up capabilities. It will be important for manufacturers of automated

30

lumber grading systems to be aware of the risks associated with bundling without

attention to customer needs.

It is hypothesized that an automated lumber grading system may have to perform

tasks other than simply assigning a grade to lumber. Depending on the production

process, machine requirements may vary. The potential benefit bundle for an automated

lumber grading system might include some or all of the following attributes:

• tallying capabilities

• sorting capabilities

• color sorting capabilities

• high through-put speed

• high durability

• simple operation

• manual override

• extended machine warranty

• extensive technical support

The importance of these attributes on machine purchase decisions among the industry

sectors mentioned earlier, furniture, cabinets, dimension, and flooring will be discussed

later.

METHODOLOGY For this research, a mail survey was used to collect data. The population of

interest included the furniture, cabinet, dimension, and flooring industries. A

31

questionnaire was initiated based on the study objectives. The questionnaire was

reviewed by members of the faculty at Virginia Polytechnic Institute and State University

for relevance to the objectives. This preliminary questionnaire was used as a script in

personal interviews with one company in each industry sector. A questionnaire was then

designed using the information obtained from the personal interviews. This questionnaire

was faxed to three companies in each sector to pretest the questionnaire before

distribution to the four sectors of the sample. Final adjustments were made using

information obtained from the pretest, and the revised questionnaire was mailed out to

1,256 companies in May 1999.

SAMPLE DEVELOPMENT

For this study, the population was the secondary wood products industry, the

sample frame consisted of the four sectors of interest; wood furniture (SIC 2511), wood

cabinets (SIC 2434), wood dimension (SIC 2426), and wood flooring (SIC 2426).

Sampling was conducted primarily from members of trade associations. It was assumed

that members of trade associations would represent the more innovative companies in the

industry; therefore, sampling from this group could introduced bias toward more

innovative companies. In addition, this study assumed that relatively innovative firms

represent the most likely adopters of an automated lumber grading system. The total

number of survey questionnaires to be mailed to each industry sector was based on an

expected 25% response rate. The most important questions in the survey addressed the

factors involved in secondary manufacturer’s adoption of an automated lumber grading

system, and sources from which they receive technological information. These questions

32

addressed the central issues outlined in the project objectives, the optimum benefit

bundle, potential adopting firms, and promotion strategy.

Ballenger and McCune (1990) state that when sampling multiple populations for

parameter studies, the following equation can be used to determine sample size. The

equation is based on a seven point rating scale (1 = least important, 7 = most important).

n = [(Zα/2)2(σ)2]/h2

where: n = sample size;

Zα/2 = reliability coefficient;

σ = estimated population standard deviation;

h = allowable tolerance level.

A confidence interval of 95% was used for this study; therefore, sample size calculations

are as follows:

Zα/2 = 1.96;

σ = (max. value – min. value)/6 = 7-1/6 = 1;

h = ± 0.2 for α = 0.05;

n = [(1.96)2(1)2]/(0.2)2 = 96.04 rounded to 97

This sample size coincides with the general consensus that a sample containing

approximately 100 units will result in significant statistical power in testing. Rossi et al.

(1983) state, as a general rule, that the sample should be sufficiently large so that there

are approximately 100 units per category. Adjusting the calculated sample size for an

expected 25% response rate, at least 388 survey questionnaires were mailed to each

sample group. Due to the limited number of wood flooring manufacturers, relative to the

33

other three sample groups, this sample was based on an attempted census of the entire

industry, 92 companies; i.e., the entire known population was surveyed. The total

number of flooring manufacturers in the US is thought to be approximately equal to the

minimum sample size calculated for this survey. A total of 1,256 survey questionnaires

were mailed to the sample frame. Actual response rates differed from the minimum

expected response rates. This is discussed further in Chapter Three.

QUESTIONNAIRE OUTLINE

One questionnaire (Appendix A) was mailed to each company in the selected

sample frame from cabinets, furniture, dimension, and flooring manufacturers. Identical

questionnaires were mailed to each sector to compare results obtained from each sector.

The questionnaire consisted of four sections; a scanning technology section, a lumber

grading section, an automated grading system section, and a demographics section. The

questionnaire was designed to engage the respondent with simple qualifying questions at

the beginning, move on to more technical questions in the body, and conclude with open-

ended questions.

Scanning questions were designed to determine which companies had scanning

technology in place, and the factors that influenced their decision to invest in this

equipment. Important factors were expected to include such aspects as increased

efficiency, accuracy, and part yield. Lumber grading questions asked for information

concerning where and when lumber was graded in the operation, and attempted to assess

the cost of manual lumber grading both monetarily and in terms of time. Questions

related to automated grading systems were designed to identify factors that would

34

influence the purchase decision for an automated grading system, and to determine key

system features influencing adoption of this technology. It was hypothesized that

important factors influencing purchase decisions might include aspects such as sorting

capabilities, tallying capabilities, technical support, and machine warranty. These

questions are important in characterizing the benefit bundle that will increase the

marketable success of automated lumber grading systems.

Finally, the demographic section was designed to help estimate company

demographics such as, company size, sales, and production levels to be used in

characterizing potential adopter firms and non-adopter firms. The questionnaire

concluded with open-ended questions allowing respondents to expand on the issues

addressed earlier in the document. Prior to analyzing the data, all responses were coded

and entered into a spreadsheet. SPSS statistical software program was used to

statistically analyze the data.

DATA ANALYSIS

The first objective, characterizing the benefit bundle that will increase the

marketable success of automated lumber grading systems, was addressed by questions in

the survey such as question 3, question 5, and question 8. Question 3 asked participants

to indicate factors that influenced the decision to purchase scanning technology in the

past, if applicable. Participants were asked to indicate the factors that influenced their

decision to invest in scanning technology as well as list any factors that did not appear on

the list. Included in the list of factors were increased efficiency, automation, part yield,

and accuracy. This type of question, resulting in categorical data, was analyzed using

35

summary statistics. Frequency tables were produced to compare the relative importance

of each factor in purchase decisions among the four sectors.

Question 5 asked participants to rate a list of factors, invoice verification,

matching grade to product, and facilitation of lumber grading on their importance in

lumber grading. The rating hierarchy was based on a 1-7 scale (1=least important,

7=most important). This type of question was analyzed using multivariate analysis of

variance (MANOVA) to test for significant differences among the four sectors; therefore,

a mean was calculated for each of the four industry sectors. The MANOVA

methodology allows for simultaneous examination of multiple samples on multiple

variables (Lehmann et al. 1998). The MANOVA technique was used to test for statistical

differences between industry sectors at a confidence level of 95%. A 95% confidence

level implies an alpha level of .05. The alpha level is the probability of type 1 error in the

statistical hypothesis, i.e., we are 95% confident that the actual mean rating value for the

population studied will, on average, be equal to the mean rating value reported in the

statistical test. Fisher’s Least Significant Difference (LSD) technique was used as a post

hoc test to detect differences between specific industry sectors.

Question 8 asked participants to rate a given set of factors on their importance in

purchasing an automated lumber grading system. Respondents were asked to rate a list

of 14 factors, on a 7-point scale (1=least important, 7=most important), on their potential

importance in a decision to invest in an automated lumber grading system. The list of

factors included: system accuracy, machine cost, technical support, equipment

warranty, reduction in labor costs, machine durability, simplicity of operation, tallying

36

capabilities, throughput speed, sorting capabilities, ability to switch species, ability to

switch grading rules, compatibility with existing equipment, and color sorting

capabilities. Question 8 was analyzed using (MANOVA) as well. MANOVA was used

to test for significant differences among the four sectors at a confidence level of 95%.

Rencher (1995) states that “MANOVA consists of a collection of methods that can be

used when several measurements are made on each individual object or object in one or

more sample.” Fisher’s LSD was used to detect significant differences between specific

industry sectors.

Objective two, identifying potential adopters and non-adopters of automated

lumber grading technology, was addressed by questions such as question 2, question 7,

and question 19. Question 2 asked the participant to indicate if they had scanning

technology in place at the time of this study. It was thought that many potential adopters

of automated lumber grading technology would have invested in other scanning

technologies in the past. For this question, frequencies were produced to assess the

percentage of each sector that had previously invested in scanning technology. Cross-

tabulations were produced to determine the relationship between technology adoption and

other demographic information.

Question 7 asked participants to estimate the number of man-hours that were

spent grading lumber, as a percentage of total man-hours. It is was thought that the

average percentage would be relatively low and that those companies who responded

with relatively high percentages may view this as added justification for investing in an

automated lumber grading system. Frequency tables were produced for question 7.

37

Question 19 related to company demographics. It asked participants to estimate

total gross sales for his/her company in 1998. It was hypothesized that there would be a

division between potential adopters and non-adopters of automated lumber grading

systems in terms of company earnings, i.e., adopters on average would consist

predominantly of companies whose gross sales were equal to or greater than some dollar

amount.

The third objective, to identify the best method of promotion to secondary

manufacturers of wood products, was addressed by questions such as question 3 and

question 14. Question 3, which also relates to the first objective, asked participants to

indicate factors that influenced their decision to invest in scanning technology in the past,

if applicable. Frequency tables were produced for data obtained from question 3. It was

thought that automated lumber grading promotional efforts should focus on those factors

that influenced investment in other scanning technologies in the past.

Question 14 asked participants to rate 9 factors on their importance to the

company in learning about new technology and industrial equipment, on a 7-point scale.

The list included: advertisements, trade shows, plant visits, scientific journals, sales calls,

meetings, trade journals, peer discussions, and unsolicited sales literature. In addition to

the 9 factors previously noted, participants were provided with an “other” category to list

factors not provided. MANOVA was used to detect significant differences among

industry sectors at a confidence level of 95%. Fisher’s LSD was used to detect

significant differences between specific industry sectors.

38

NON-RESPONSE BIAS

This study used a sample frame consisting of cabinet, furniture, dimension, and

flooring manufacturers from across the US; however, not all manufacturers receiving a

questionnaire participated in the survey. To be sure that the sample was truly

representative of the population, comparisons were made between respondents and non-

respondents. Ten companies in each industry sector were contacted by phone and asked

a series of questions selected from the questionnaire. Questions from which means could

be calculated were chosen for the comparison.

Questions 8 and 17 were selected. Question 8 consists of fourteen system

attributes for which respondents were asked to rate their importance in the adoption of

automated lumber grading technology on a 1-7 scale (1=least important, 7=most

important). For brevity only five attributes were used in the phone survey. These

attributes included system accuracy, machine cost, technical support, tallying capability,

and sorting capability. Question 17 asked participants to estimate the company’s annual

lumber input volume in board feet. ANOVA was used to compare means between

respondents and non-respondents on questions 8 and 17 at a confidence level of 95%. No

significant differences were detected on the technical support, tallying capabilities,

sorting capabilities, or annual lumber input variables. However, significant differences

were detected on system accuracy and machine cost. No significant differences were

detected among industry sectors on annual lumber input volume.

Another method often used to test how well the sample represents the population

studied is to compare early respondents to late respondents. It is thought that late

39

responses are a valid representation of non-responses. A comparison was performed

between the first and last thirty responses received using a t-test of the means. The t-test

was performed at a 95% confidence level. Significant differences were not detected for

machine cost, technical support, tallying capability, sorting capability, or estimated

annual lumber input. However, a significant difference was detected between early and

late respondents on the machine accuracy attribute. It was reassuring to find that

significant differences were not detected between, neither respondents versus non-

respondents, nor early versus late respondents on most variables. However, the presence

of significant differences on system accuracy in both comparisons indicates that our

findings on this attribute may not be representative of the population.

PROJECT SUMMARY

The overall goal of this study was to assess, not quantify, market potential for an

automated lumber grading system in the secondary wood processing industry. This study

was designed to determine why a secondary wood processing company would or would

not invest in automated lumber grading technology.

A mail survey was used to collect data from companies in the four sectors,

furniture, cabinets, dimension, and flooring manufacturers. The questionnaire was

designed to meet the objectives of this project. The objectives were as follows: (1) to

characterize potential adopters and non-adopters of automated lumber grading

technology, (2) to characterize the benefit bundle that will increased the marketable

success of an automated lumber grading system in the secondary wood processing

40

industry, and (3) to identify the best methods of promotion to the secondary wood

processing industry.

This study was designed to assist manufacturers of wood scanning technologies in

understanding product attributes that are important to the secondary wood processing

industry and differences in those attributes among the four sectors. In addition, this study

was designed to help manufacturers of wood scanning technologies with product

planning and development. This study will assist manufacturers of wood scanning

technologies in targeting potential markets and identifying those companies that are more

likely to invest in this type of equipment. Finally, this study identified important product

attributes on which to focus in promotional efforts, and the most beneficial methods of

promotion to the secondary wood processing industry.

Chapter Two will address objective one, differences between adopters versus non-

adopters of scanning technology, and discuss overall market demographic information.

Chapter Three will address objectives two and three, characterization of the benefit

bundle that will increase the overall marketable success of automated lumber grading

technology, and identification of the best methods of promotion to the secondary wood

products industry. Chapter Four will summarize this research and conclusions relative to

the objectives and prescribe recommendations for future research.

41

REFERENCES Adams, Larry. 1999. Cabinet Companies Step Out of the Kitchen. Wood and Wood Products. Vol. 104, No. 3, pages 51-63. Anonymous. 1999. National Hardwood Lumber Association. http://www.natlhardwood.org/main.html (24 Jan. 1999) Anonymous. 1999. National Wood Flooring Association. http://www.woodfloors.org/species.shtml. (2 Feb. 1999) Anonymous. 1999. Process Automation. http://www.vtwood.forprod.vt.edu (12 Feb. 1999). Anonymous. 1977. Defect Analysis May Boost Lumber Yield by 15%. Forest Industries. Vol. 104, No. 1, 1 page. Ballenger, Joe K. and S. McCune. 1990. A Procedure for Determining Sample Size for Sampling Multiple Populations for Parameter Studies. Journal of Marketing Education. Vol. 12, No. 3, pages 30-33. Capon, Noel and R. Glazer. 1987. Marketing and Technology: A Strategic Coalition. Journal of Marketing. Vol.51, No. 7, pages 1-14. Choffray, Jean-Marie, and G. L. Lilien, 1980. Market Planning for New Industrial Products. John Wiley and Sons. New York. 294 pages. Christianson, Rich. 1998. Survey Sneaks a Peak at 2000 Trends. Wood & Wood Products. December, 1998, pages 44-47. Conners, Richard W., D. E. Kline, P. A. Araman, X. Xiangyu, T. H. Drayer, inventors; Virginia Polytechnic Institute and State University, assignee. Automated lumber grading system. US patent 1996000707982. 1999, September 28. Cooper, Robert G. 1994. New Products: The Factors that Drive Success. International Marketing Review. Vol. 11, No.1, pages 60-76. Cooper, R. G. 1979. The Dimensions of Industrial New Product Success and Failure. Journal of Marketing. Vol. 43, Summer, pages 93-103. Davidow, William H. 1986 Marketing High Technology, An Insider’s View. The Free Press. New York, pages 133-135.

42

Dempsey, Gilbert P. and W. G. Luppold. 1991. The State of Hardwood Lumber Markets. Northern Logger and Timber Processor. March, 3 pages. Doud, Lynn F., W. W. Johnson, P.R. Blankenhorn, L. R. Stover, and W. G. Luppold. 1992. A Description of the 1987 Kitchen Cabinet Industry in Pennsylvania. Forest Products Journal. Vol. 42, No. 4, pages 33-39. Hansen, Bruce and C. West. 1998. Trends in Domestic Hardwood Markets. Hardwood Symposium Proceedings. May 6-9, 9 pages. Hansen, Eric, V. S. Reddy, J. Punches, and R. Bush. 1995. Wood Material Use in the US Furniture Industry: 1993 and 1995. Center for Forest Products Marketing, Virginia Polytechnic Institute and State University. 19 pages. Hass, Michael, P. and P. M. Smith. 1997. Global Markets for US Hardwood Components. Forest Products Journal. Vol. 47, No. 3, pages 45-51. Hoff, Kristen, N. Fisher, S. Miller, and A. Webb. 1997. Sources of Competitiveness for Secondary Wood Products Firms: A Review of Literature and Research Issues. Forest Products Journal. Vol. 47, No. 2, pages 31-37. Huber, Henry A., C. W. McMillin, and J. P. McKinney. 1985. Lumber Defect Detection Abilities of Furniture Rough Mill Employees. Forest Products Journal. Vol. 35, No. 11/12, pages 79-82. Hutt, Michael D. and T. W. Speh. 1981. Industrial Marketing Management. The Dryden Press. Chicago, 463 pages. Kline, D. E., Connors R. W., and Araman P. A. 1998. What’s Ahead in Automated Lumber Grading, Proceedings of Scan-Pro Conference. Vancouver, British Columbia, Canada, 11 pages. Lehman, Donald R. and J. O’Shaughnessy. 1974. Difference in Attribute Importance for Different Industrial Products. Journal of Marketing. Vol. 38, No. 4, pages 36-42. Lehman, Donald R., S. Gupta, and J. H. Steckel. 1998. Marketing Research. Addison-Wesley. Reading, Mass., 780 pages. Mittal, Vikas, W. T. Ross, Jr., and P. M. Baldasare. 1998. The Asymmetric Impact of Negative and Positive Attribute-Level Performance on Overall Satisfaction and Repurchase Intentions. Journal of Marketing. Vol. 62, No. 1, pages 33-47. More, Roger A. 1986. Developer/Adopter Relationships in New Industrial Product Situations. Journal of Business Research. Vol. 14, pages 501-517.

43

Mullin, Sandy. 1998. Scanning Lumbr for Defects and Increased Profit. Wood & Wood Products. February, 1998, pages 37-38. Ostlund, Lyman E. 1974. Perceived Innovation Attributes as Predictors of Innovativeness. Journal of Consumer Research. Vol. 1, No. 9, pages 23-29. Patterson, David W., Anderson, and Rockwell. 1997. Increased Use of Automated Machinery Requires Changes in Quality Control Procedures. Forest Products Journal. Vol. 47, No. 1, pages 33-36. Pham, Duc Truong and R. J. Alcock. 1998. Automated Grading and Defect Detection: A Review. Forest Products Journal. Vol. 48, No. 4, pages 34-42. Porter, Michael E. 1980. Competitive Advantage. The Free Press. New York, pages 425-442. Punches, John W., E. N. Hansen, and R. J. Bush. 1995. Productivity Characteristics of the US Wood Cabinet Industry. Forest Products Journal. Vol. 45, No. 10, pages 33-38. Rencher, Alvin C. 1995. Methods of Multivariate Analysis. John Wiley & Sons. New York, NY. 626 pages Rhyne, Lawrence C. 1994. Product Development with a New Technology. Journal of Product and Brand Management. Vol. 3, No. 3, pages 39-50. Rosenburg, Nathan, P. Ince, K. Scog, and A. Plantinga. 1990. Understanding the Adoption of New Technology in the Forest Products Industry. Forest Products Journal. Vol. 40, No. 10, pages 15-22. Rossi, Peter H., J. D. Wright, and A. B. Anderson. 1983. Handbook of Survey Research. Academic Press. New York, page 157. Saunderson, Allan. 1982. Computers Keep Lifeline Afloat. Canadian Forest Industries. August, 2 pages. Weinig, Michael. 1999. Parquet Flooring An Expanding Market. Wood and Wood Products. Vol. 104, No. 2, pages 89-90. Williston, Ed. 1979. Lumber Manufacture (State of the Art). Forest Products Journal. Vol. 29, No. 10, pages 45-49. White, Vernon S. 1977. First Edger-Optimizer Built in US Goes in Small Log Mill. Forest Industries. Vol. 104, No. 1, 1 page.

44

Vlosky, Richard P. 1996. Characteristics of US Hardwood Wood Component Manufacturers. Forest Products Journal. Vol. 46, No. 5, pages 37-43.

45

CHAPTER 2

ABSTRACT

In determining potential markets for scanning technology, specifically an

automated lumber grading system (AGS), it is important to understand demographic

differences that exist between segments of the target market. For this research scanning

technology is defined as process optimization technology, which uses various sensing

technologies coupled with computer memory to perform such tasks as defect detection,

optimized sawing, grading, sorting, and tallying of materials. Four groups were

evaluated in this study, cabinets, furniture, dimension, and flooring manufacturers to

characterize the market for lumber scanning technology in the secondary wood products

industry, and identify differences between adopters and non-adopters of automated

lumber grading technology. The cabinet and furniture samples in this study were

predominantly comprised of smaller companies. This information is important to

manufacturers of scanning technology in marketing their products and to the wood

products industry in receiving information concerning this technology relative to their

individual businesses. This study found that potential adopters of automated lumber

grading technology will most likely be larger companies, both in terms of sales and

number of employees. The majority of adopting firms reported annual sales in excess of

$5 million and employed greater than 100–200 people. In addition, analysis showed that

dimension and flooring manufacturers may better represent adopters of automated lumber

grading technology due to higher lumber throughput volumes and higher lumber grading

costs.

46

INTRODUCTION

The secondary wood products industry represents a potential market for scanning

technology, due to the high volumes of hardwood lumber throughput that are

characteristic of companies in this industry. For this research scanning technology is

defined as process optimization technology, which uses various sensing technologies in

conjunction with computer memory to perform such tasks as defect detection, optimized

sawing, lumber grading, and sorting and tallying of materials. The secondary wood

products industry produces a wide range of products, which require similar but different

processing techniques. Products produced by this industry range from railway ties to

bedroom furniture. In addition, due to increased competition, changes in policies

influencing the availability of raw materials, and a changing global economy, companies

in this industry are searching for ways to improve their production processes and remain

competitive. Scanning technologies, used in other industries for years, coupled with

computer technology, can help secondary wood products manufacturers achieve their

goals of improving production efficiency and part yield, among other manufacturing

concerns. The goal of this study was to characterize potential adopters and non-adopters

of lumber scanning technology, specifically automated lumber grading technology. This

information is necessary in identifying the target market for this technology in the

secondary wood products industry.

The secondary wood products industry is a sizeable industry in terms of hardwood

lumber consumption. Greater than 90% of all hardwood lumber produced in the US is

consumed by eight secondary wood processing markets. These markets include pallets,

47

furniture, dimension parts, exports, millwork, flooring, cabinets, and railway ties.

Hardwood lumber consumption for these markets has been estimated at over 13 billion

board feet in 1997 (Hansen and West 1998). The four sectors included in this study,

furniture, cabinets, dimension, and flooring, consumed approximately 52% of all

hardwood lumber produced in the US in 1997, an estimated 7.7 billion board feet. In

addition, all four sectors showed an increase in hardwood lumber consumption in this

decade (Hansen and West 1998). US consumption figures for 1997 show a total

hardwood lumber consumption increase of over 2 billion board feet since 1991 (Dempsey

and Luppold 1992).

While the wood products industry has not been traditionally known for its

utilization of high-technology equipment, many within the industry are seeking methods

to automate their processes to increase part yield and product value. In addition, many

companies are searching for methods to regulate the precision of their processes, improve

their quality control operations, and facilitate the tallying and sorting of materials.

Patterson et al. (1997) noted that 36% of furniture manufacturers and 55% of cabinet

companies had some type of computer numeric control (CNC) machinery in place. In

addition, greater than 39% of the furniture companies and 13% of the cabinet

manufacturers indicated that they planned to purchase a CNC machine within 12 to 24

months.

An area of automation that is currently diffusing into the wood products industry

is scanning technology, which uses various types of sensors to scan lumber and computer

memory to gather and interpret data from the scanning apparatus. Technologies such as

48

camera, laser, and x-ray represent some of the technology available in scanning systems

today. The information gathered by these scanning systems can be used in defect

detection, automated lumber grading, and sawing optimization. Lumber scanning

technology for grading and processing hardwood lumber is now in the development

stages. However, it is expected to become more commonplace in the industry as the

proficiency of the technology advances and is proven to the industry.

“Most policy makers believe adoption of advanced technologies is fundamental to

wood manufacturers’ competitiveness in global markets” (Hoff et al. 1997). Current

justification for purchasing a scanning system is the reduction of labor. However, the

value of information that can be received from scanners may be of more importance.

Kline et al. (1998) state that “The primary cost savings from an automated lumber

grading system will be realized by producing a more uniform and consistently graded

product and by producing a higher value product through optimum lumber

remanufacturing.” The ability to optimize yield and value are perhaps the most

important attributes of lumber scanning systems, however, these systems are limited by

the information they receive (Mullin 1998). In considering the impact on yield using

automated lumber grading technology, it is important to understand that any increase in

yield will not be directly related to more accurate lumber grading. It is thought that an

automated lumber grading system may contribute to increased yield by providing the

optimum cut-up algorithm through accurate defect mapping. This implies that the

algorithm generated by an automated grading system is only as reliable as the

information received from the scanning heads. In addition, scanning systems are

49

currently new industrial products, which may require different marketing strategies

compared to new household consumer products.

“Marketing new products to industrial firms remains a mystery to many

manufacturers” (Berkowitz 1986). Different firms take different approaches to

marketing their products to industrial end-users. Some production-oriented firms believe

that if they develop a superior product and get it into the hands of the end-users then sales

will naturally follow. Some sales-oriented firms believe that they can gain market share

by simply offering a lower price than their competitors, while other firms believe that an

expensive advertising campaign is necessary. The truth may be, however, that there is a

complicated process involved in marketing industrial products (Berkowitz 1986). This

process involves such tasks as identification of a need, market analysis, product

development and testing, and sales. Of particular importance is the identification of

potential adopters of the product. By identifying demographic characteristics of potential

adopters, firms can avoid targeting the wrong markets for their products.

It is obvious that the adoption of lumber scanning technology by the secondary

wood products industry depends largely on the performance of these systems in actual

industrial applications. Rosenberg et al. (1990) noted that technology adoption in the

forest products industry is influenced by the role of innovations imported from other

industries, the effect of raw material shortages, the importance of economic factors in

adoption of innovations, and problems associated with the heterogeneity of wood raw

material and finished products.

50

Technology is often developed at the expense of one industry, while other

industries later enjoy the benefits of the technology. For example, x-ray technology has

been used in the medical field for a number of years, and in the case of automated lumber

grading systems, it is now diffusing into the forest products industry. In addition, wood

products manufacturers will focus more on improving processing operations in periods of

economic recession and at times when raw material quality is not consistent. These

scenarios lead to greater emphasis on increasing product value and part yield, and result

in more innovation within the industry.

OBJECTIVES

The goal of this study was:

• to characterize the market segments that may utilize hardwood scanning technology.

The objective of this study was:

• to identify differences which may exist between potential adopters and non-adopters

of hardwood scanning technology.

METHODOLOGY

A questionnaire was initiated based on the study objectives. The questionnaire

was then reviewed by faculty at Virginia Polytechnic Institute and State University for its

content. The questionnaire was developed by personally interviewing persons from one

company in each industry sector. These interviews were used to identify important

information that should be included in the questionnaire. Further adjustments were then

51

made to the questionnaire and it was pre-tested by fax with three companies in each

industry sector. The questionnaire was pre-tested to ensure that questions were clear and

followed a logical format. Participants were asked to return the completed survey along

with any suggested changes. Final adjustments were made following the pre-test, and the

revised questionnaire was mailed in May 1999 to the sample frame which included 1,256

companies.

The sample frame consisted of four sectors of the secondary wood products

industry; cabinets, furniture, dimension, and flooring manufacturers. The sample frame

was predominantly comprised of members of various trade associations. It was assumed

that members of trade associations would represent companies who would be more likely

to adopt new technology. This sample frame may be biased toward more innovative

companies. The study objectives sought to identify companies who would most likely

become adopters of automated lumber grading technology.

RESULTS

Data was entered into SPSS statistical software package for analysis when

received. A total of 439 responses were received. Of those, 209 were useable responses

resulting in an overall response rate of 16.6%. Adjusting the overall response rate for

undeliverable questionnaires (175), and companies performing operations not applicable

to the sample (55), the overall adjusted response rate was calculated at 20.4%. The

numbers of questionnaires received according to market segment were: cabinets (45),

furniture (35), dimension (92), and flooring (37). These responses resulted in response

52

rates of 11.6% for cabinet manufacturers, 9% for furniture manufacturers, 23.7% for

dimension manufacturers, and 40.2% for flooring manufacturers. Adjusting the response

rates per industry segment resulted in 13.6% for cabinet manufacturers, 12.4% for

furniture manufacturers, 28.4% for dimension manufacturers, and 45.1% for flooring

manufacturers.

MARKET CHARACTERIZATION

Questions were included in the study to be used in comparing the relative size of

each industry sector in terms of such company characteristics as sales volume, lumber

throughput, and number of employees. In addition, this type of information was used to

understand the relationship of demographic information to adoption of new technology.

The sample mean for annual lumber throughput for all manufacturers was calculated at

7.3 million board feet. Mean values for annual lumber throughput, given in Figure 2,

were calculated at 1.3 million board feet for cabinet manufacturers, 2 million board feet

for furniture manufacturers, 7.7 million board feet for dimension manufacturers, and 18.3

million board feet for flooring manufacturers.

53

Mean Lumber Throughput by Business Type

0

5,000,000

10,000,000

15,000,000

20,000,000

cabinet furniture dimension flooring

Business Type

Boa

rd F

eet

Figure 2: Mean lumber production volumbe according to industry group

Table 2 shows that of the four industry sectors in the study, greater than 41% of

the cabinets, dimension, and flooring sectors responded with annual sales in excess of

$10 million. The majority of furniture manufacturers who responded, 55.2%, reported

annual sales in the range of $1 million to $5 million. In addition, the cabinet sector

reported the largest percentage, 22%, of companies with annual sales less than $500,000.

Among respondents from the four industry sectors, all reported a majority of

companies employing fewer than 200 people (Table 2). The cabinet, furniture, and

dimension sectors reported a majority of companies employing fewer than 100 people. In

addition, the cabinet and flooring sectors reported the largest percentage of respondents

54

with greater than 400 employees at 15% and 13.5% respectively. This information

indicates that the sample frame represents smaller cabinet and furniture manufacturers.

Table 2: Percentage of sales and employees per industry sector

Annual Sales 1998 Cabinets Furniture Dimension Flooring Overall Less than $100,000 0 0 2.3% 0 1.1%$100,000 - $500,000 22% 6.9% 2.3% 0 7.9%$500,001 - $1,000,000 7.3% 0 4.7% 3% 4.2%$1,000,000 - $5,000,000 24.4% 55.2% 30.2% 15.2% 30.2%$5,000,001 - $10,000,000 4.9% 13.8% 18.6% 15.2% 14.3%Greater than $10,000,000 41.5% 24.1% 41.9% 66.7% 43%Full-time Employees OverallFewer than 25 37.5% 21.9% 21.1% 10.8% 22.6%25 - 50 20% 25% 23.3% 13.5% 21.1%51 - 100 2.5% 21.9% 18.9% 21.6% 16.6%101 - 200 7.5% 12.5% 22.2% 29.7% 19.1%201 - 300 10% 6.3% 7.8% 8.1% 8%301 - 400 7.5% 3.1% 1.1% 2.7% 3%Greater than 400 15% 9.4% 5.6% 13.5% 9.5%

Response Frequency

Respondents from the cabinet and furniture sectors reported that the majority of

both industry sectors did not market their products internationally, 83.8% and 68.8%

respectively. In contrast, the dimension and flooring sectors reported a majority of

respondents selling products in foreign markets, 56% and 67.6% respectively (Table 3).

In addition, respondents from all sectors reported that greater than 84.5% of companies in

each sector was privately owned. The cabinet sector showed the largest percentage of

publicly owned companies at 15.5%.

Participants were asked to indicate if they had any scanning technology in place at

the time of this study. Table 3 shows all sectors reporting a majority of companies

answering no to this question. Eighty-one percent of the cabinet sector did not have

scanning technology in place at the time of this study, while 85.7% of the furniture sector

did not have scanning technology in their plants. Percentages of companies in the

55

cabinets and furniture sectors that had invested in scanning technology were 19% and

14.3% respectively. These findings may be influenced by the relative smaller sizes of

cabinet and furniture manufacturers in the sample frame. Percentages reported by the

dimension and flooring sectors were similar as well. Greater than 61% of the dimension

sector and almost 63% of the flooring sector did not have scanning technology in place at

the time of this study, while 38.5% and 37.1% of the dimension and flooring sectors

respectively had invested in scanning technology.

Table 3: Percentage of international marketing, company ownership, scanning tech adoption according to sector

Cabinets Furniture Dimension Flooring OverallNo 83.8% 68.8% 44% 32.4% 53.5%Yes 16.2% 31.3% 56% 67.6% 46.5%

Company Ownership OverallPublic 15.5% 9.4% 7.7% 8.1% 9.4%Private 84.5% 90.6% 92.3% 91.9% 90.6%

Have Scanning Tech OverallNo 81.0% 85.7% 61.5% 62.9% 70%Yes 19.0% 14.3% 38.5% 37.1% 30%

International MarketingResponse Frequency

A lumber grading section was included in the questionnaire to assess the cost of

grading lumber both in monetary costs and in terms of time. A mean of $14.66 per man-

hour was calculated for the entire sample frame. Figure 3 shows a mean cost of $12.11

per man-hour was calculated for the cabinet sector, while the furniture sector reported a

mean cost of $12.31 per man-hour. The dimension and flooring sectors reported higher

lumber grading costs than the cabinet and furniture sectors. Mean lumber grading costs

were calculated at $18.21 and $16.01 per man-hour for the dimension and flooring

sectors respectively.

56

Cost Per Man-hour of Grading Lumber by Group

$0.00$5.00

$10.00$15.00$20.00

cabin

et

furnit

ure

dimen

sion

floori

ng

Business Type

US

Dol

lars

Figure 3: Mean man-hour cost of lumber grading according to industry group

Participants were asked to indicate how many man-hours were spent in their

plants grading lumber as a percentage of total production man-hours. Table 4 illustrates

the cabinets and furniture sectors reporting a majority of companies in both sectors

dedicating less than 2% of total man-hours to grading lumber. Percentages of companies

in the cabinets and furniture sectors utilizing less than 2% of total man-hours to grade

lumber were 81.3% and 85.2% respectively. In comparison, the dimension and flooring

sectors showed more companies spending more time grading lumber. Percentages of

companies in the dimension and flooring sectors dedicating less than 2% of man-hours to

lumber grading were lower at 27.6% and 50% for dimension and flooring respectively.

In addition, 16.1% of dimension manufacturers and 15.6% of flooring manufacturers

57

indicated that greater than 10% of total man-hours in their plants were dedicated to

lumber grading, while the cabinets and furniture sectors reported no companies spending

greater than 10% of man-hours grading lumber.

Participants were also asked to indicate what their expected payback period would

be for the purchase of an automated lumber grading system. A range of less than two

years to greater than ten years was provided. The majority of the four sectors in the study

responded with expected payback periods of less than six years (Table 4). Greater than

85% of cabinets manufacturers, 60% of furniture manufacturers, 87.9% of dimension

manufacturers, and 80.6% of flooring manufacturers responded with a payback period of

less than six years. The cabinet sector reported the highest percentage of respondents

with expected payback periods of less than two years at 48.1%. In addition, the furniture

sector reported the highest percentage of respondents with expected payback periods

greater than ten years, 24%.

Table 4: Percentage of man-hour costs of grading lumber, payback period for automated lumber grading system according to industry group

Man-hours Grading Lumber Cabinets Furniture Dimension Flooring OverallLess than 2% 81.3% 85.2% 27.6% 50.0% 50%3% - 5% 12.5% 11.1% 39.1% 18.8% 26.4%6% - 9% 6.3% 3.7% 17.2% 15.6% 12.9%Greater than 10% 0 0 16.1% 15.6% 10.7%Expeted Payback for AGS OverallLess than 2 years 48.1% 20% 34.9% 30.6% 33.9%3 - 5 years 37% 40% 53% 50% 48%6 - 10 years 3.7% 16% 4.8% 8.3% 7%Greater than 10 years 11.1% 24% 7.2% 11.1% 11.1%

Response Frequency

Analysis of overall response shows that the dimension and flooring sectors

process more lumber volume, on a board foot basis, than the cabinet and furniture

sectors. However, demographic information indicates that our cabinet and furniture

58

industry samples were comprised of predominantly smaller companies. This explains the

relatively low, mean lumber throughput values reported among companies in the cabinet

and furniture sectors. Furthermore, dimension and flooring manufacturers incurred

higher lumber grading costs compared to cabinets and furniture manufacturers and were

found to spend more time grading lumber, as a percentage of total man-hours. The

flooring sector reported the largest percentage of companies with annual sales in excess

of $10 million. It was also found that more dimension and flooring manufacturers market

their products internationally compared to cabinet and furniture manufacturers, and a

large majority of study participants were privately owned. In addition, larger percentages

of the dimension and flooring sectors had adopted lumber scanning technology than the

cabinet and furniture sectors. Again, these findings may be influenced by the relatively

small companies, which comprised the majority of both the cabinet and furniture

samples. Overall response information implies that the market for automated lumber

grading technology, in the secondary wood products industry, may be concentrated in the

dimension and flooring sectors. Overall demographic information is useful in identifying

adopters versus non-adopters of hardwood scanning technology. This information can be

used to characterize a company profile for potential adopters of scanning technology.

DIFFERENCES AMONG ADOPTERS AND NON-ADOPTERS

Identifying potential adopters and non-adopters is important to manufacturers of

scanning technology in determining appropriate markets for their products. This study

identified differences between potential adopters and non-adopters of scanning

technology in the secondary wood products industry. The study focused specifically on

59

automated lumber grading technology, in the cabinet, furniture, dimension, and flooring

industries.

The study determined if participants believed that an automated lumber grading

system would benefit their business. A clear distinction was found to exist between the

cabinet and furniture sectors and the dimension and flooring sectors. Table 5 summarizes

the cabinet and furniture sectors 65.7% and 78.3%, respectively, stating that an

automated lumber grading system would be of no benefit to them. The dimension and

flooring sectors reported 67.1% and 72.4% respectively responding positively to the

benefits of automated lumber grading technology. This was supported by the

demographic findings that dimension and flooring manufacturers spend both more time

and money grading lumber in their operations. An explanation of differences in

perception of the benefit of automated lumber grading technology may be that these

differences are related to the variation in production processes among these sectors. In

other words, it may be more difficult for cabinet and furniture manufacturers to see the

benefit of automated lumber grading technology due to their longer production chains

compared to dimension and flooring manufacturers. In the dimension and flooring

industries the entire production process is concentrated in the rough mill, while the rough

mill represents only the beginning of the process in cabinet and furniture production. In

addition, it should be noted that our cabinet and furniture samples were predominantly

comprised of smaller companies. The concentration of smaller companies in these

samples indicates that our findings are representative of a subset of the cabinet and

furniture industries.

60

Table 5: Percentage of companies finding automated lumber grading system beneficial, companies contacted by competitors concerning new equipment according to industry sector

AGS Beneficial Cabinets Furniture Dimension Flooring OverallNo 65.7% 78.3% 32.9% 27.6% 45.2%Yes 34.3% 21.7% 67.1% 72.4% 54.8%

Contacted OverallNo 52.8% 55.6% 39.8% 45.9% 45.7%Yes 47.2% 44.4% 60.2% 54.1% 54.3%

The study determined if companies in each industry sector contacted other

companies in their industry to inquire about high-technology equipment. Participants

were asked to indicate whether they were or were not contacted by other companies.

This information was used to help understand how technology information is shared

within each industry sector. In addition, it was used to determine if there were companies

in the sample frame that were considered to be sufficiently innovative such that other

companies would contact them about high-tech equipment they were considering. It was

believed that this would be a factor in determining perceived innovativeness of a

company. Table 5 illustrates the furniture and cabinet sectors reported nearly equivalent

percentages. Greater than 52% of the cabinet sector responded that they had not been

contacted concerning high-tech equipment they were considering, while 55.6% of the

furniture sector reported that they had not been contacted. In contrast, the dimension and

flooring sectors reported that 60.2% and 54.1% of companies respectively had been

contacted concerning high-tech equipment. This finding implies that information is

shared more freely among companies in the dimension and flooring sectors.

The previous discussion indicates that cabinet and furniture manufacturers find

little benefit in automated lumber grading compared to dimension and flooring

manufacturers, and that dimension and flooring manufacturers are more likely to share

61

information concerning new processing technologies with competitors. The following

discussion will explore more specific relationships influencing these findings. This

discussion includes demographic relationships between likely adopters and non-adopters.

Participants were asked to estimate their average cost per man-hour of grading lumber.

Mean costs of $12.11, $12.31, $18.21, and $16.01 per man-hour were found for the

cabinets, furniture, dimension, and flooring groups respectively (Figure 3). Analysis of

variance (ANOVA) was used to compare mean values to determine if significant

differences existed among the industry sectors (Table 6). Fisher’s LSD technique was

used as a post hoc test to determine between which sectors differences had occurred. A

significance level of .01 was calculated indicating that a difference did exist among the

sectors. Significant differences were detected between the cabinet and dimension sectors

and the furniture and dimension sectors. Significance levels for these comparisons were

less than .01 for both comparisons. This indicates that the dimension sector on average

incurred higher lumber grading costs, on a man-hour basis, than the cabinet and furniture

sectors among the companies sampled in this study. This information is consistent with

demographic findings that the dimension and flooring sectors, on average, processed

more lumber than the cabinet and furniture sectors among the companies sampled in this

study.

Table 6: Mean man-hour cost of grading lumber according to industry sector Significance

Cabinets (1) Furniture (2) Dimension (3) Flooring (4) LevelMean Cost Grading Lumber 12.11 12.31 18.211 2 16.01 0.01*

Significance level: cabinets and dimension comparison < .01Significance level: furniture and dimension comparison < .01

*Represents significant differences among groups using analysis of variance1 2 Represents significant differences between specific groups

62

In addition to greater lumber grading costs in the dimension and flooring

industries, demographic information shows these sectors overall utilizing more scanning

technologies than the cabinets and furniture sectors. It was noted in Table 3 that the

majority of all four sectors did not have scanning technology in place at the time of this

study, however the dimension and flooring sectors reported more companies using

scanning technology than the cabinets and furniture sectors. Percentages of companies in

the sample using scanning technology were 38.5% for the dimension sector and 37.1%

for the flooring sector compared to 19% for the cabinets sector and 14.3% for the

furniture sector.

In addition, responses were used in cross-tabulations to show the relation of

scanning technology adoption to demographic information such as company sales,

international marketing activity, and number of employees (Table 7). It was thought that

this information would help in developing a profile of potential adopters of automated

lumber grading technology. While the majority of respondents did not have scanning

technology, no company with less than $500,000 in annual sales had invested and only

one company with less than $1 million in annual sales had invested in scanning

technology. Of those companies who had scanning technology in place, 89% reported

greater than $5 million in annual sales. Of those companies who had not invested in

scanning technology, only 42.5% reported greater than $5 million in annual sales. This

implies that adopters of scanning technology will most likely be large companies in terms

of annual sales.

63

Furthermore, Table 7 illustrates the relationship between adoption of scanning

technology and international product marketing activity, it was noted that approximately

50% of those companies who had invested in scanning technology also marketed their

products internationally. Of those companies who did not have scanning technology in

place, approximately 20% sold their products in foreign markets. International marketing

activity is often perceived as a characteristic of innovative firms. Therefore, by this logic

it is implied that adopters of scanning technology will most likely be perceived as more

innovative companies within the industry.

Using cross-tabulation to show the relationship between the adoption of scanning

technology and number of employees, it was noted that 61.7% of those companies who

had invested in scanning technology employed more than 100 people (Table 7). In

addition, of those companies who did not have scanning technology in place, only 30.1%

employed more than 100 people. This implies that adopters of scanning technology will

most likely be large companies in terms of number of employees.

Table 7 also illustrates the relation of adoption of scanning technology to

expected cost of an automated lumber grading system. It was noted that the majority of

companies who had invested in scanning technology expected higher machine costs than

companies who had not invested in scanning technology. Greater than 59% of those

companies who had scanning technology in place expected an automated lumber grading

system to cost in excess of $150,000. In comparison, approximately 33% of companies

who did not have scanning technology in place expected an automated lumber grading

system to cost greater than $150,000. This indicates that companies who have previously

64

invested in scanning technology may expect higher costs associated with hardwood

scanning systems than those who have not invested in scanning technology.

Participants were asked to rate a list of system attributes in terms of their

importance in investing in automated lumber grading technology. The rating scheme was

based on a (1-7) rating scale, where (1=least important, 7=most important). The list of

attributes included: (machine accuracy, cost, durability, compatibility with existing

equipment, throughput speed, ability to quickly switch species, ability to quickly switch

grading rules, simplicity of operation, technical support, equipment warranty, reduction

in labor costs, tallying capabilities, sorting capabilities, and color sorting capabilities).

Analysis of variance (ANOVA) was used to test for significant differences between the

two groups, past adopters and non-adopters, at a confidence level of 95%.

65

Table 7:Relation of adoption of scanning technology to company demographic information

T o ta l S a le s in 1 9 98 N o Y e s T o ta lL e ss th a n $1 0 0 ,0 00 2 2$ 1 00 ,0 0 0 - $ 5 0 0 ,0 0 0 1 3 1 3$ 5 00 ,0 0 1 - $ 1 ,0 00 ,0 0 0 7 1 8$ 1 ,00 0 ,0 01 - $ 5 ,0 0 0 ,0 00 5 1 5 5 6$ 5 ,00 0 ,0 01 - $ 1 0 ,0 0 0 ,00 0 1 4 1 1 2 5$ G rea te r tha n $ 1 0 ,0 0 0 ,00 0 4 0 3 9 7 9T o ta l 1 2 7 5 6 1 8 3

M a rk e t In te rn a tio n a lly N o Y e s T o ta lN o 8 5 4 8 1 3 4

Y e s 2 0 4 0 6 0T o ta l 1 0 5 8 8 1 9 4

N u m b e r o f E m p lo y ee s N o Y e s T o ta lF e w e r th a n 2 5 4 2 3 4 52 5 - 5 0 3 1 9 4 05 1 - 1 0 0 2 0 1 1 3 11 0 1 - 2 0 0 2 1 1 6 3 72 0 1 - 3 0 0 8 8 1 63 0 1 - 4 0 0 3 3 6G re a te r th an 4 0 0 8 1 0 1 8T o ta l 1 3 3 6 0 1 9 3

E x p ec te d C o s t o f A G S N o Y e s T o ta lL e ss th a n $5 0 ,0 0 0 2 9 4 3 3$ 5 0 ,0 0 0 - $ 1 5 0 ,00 0 4 3 1 6 5 9$ 1 50 ,0 0 1 - $ 2 5 0 ,0 0 0 1 9 1 1 3 0$ 2 50 ,0 0 1 - $ 5 0 0 ,0 0 0 1 0 1 3 2 3G re a te r th an $ 5 0 0 ,0 0 0 6 5 1 1T o ta l 1 0 7 4 9 1 5 6

H a ve s c an n in g te ch

H a ve s c an n in g te ch

H a ve s c an n in g te ch

H a ve s c an n in g te ch

Table 8 illustrates that overall mean values exceeded the average value of (4) for

all listed factors, therefore the listed factors were considered to be at least somewhat

important. In addition, machine accuracy was rated overall highest at 6.08, while the

color sorting capabilities factor was rated the overall lowest at 4.35. Significant

differences were detected between adopters and non-adopters of scanning technology on

throughput speed, tallying capabilities, and sorting capabilities. Significance levels were

66

.03 for the throughput speed factor, .03 for the tallying capabilities factor, and .01 for the

sorting capabilities factor. In all cases companies who had invested in scanning

technology rated these factors higher than those who had not invested in scanning

technology. This indicates that these factors were more important to past adopters of

scanning technology in their decision to invest in automated lumber grading technology

compared to companies without scanning technology in their operations. This may be an

indication of characteristics of potential adopter firms, i.e., potential adopters of

automated lumber grading technology may operate at higher production speeds, and have

a greater need for enhanced lumber tallying and sorting capabilities.

Table 8: Comparison of past adopters vs. non-adopters on system attributes

System Overall SignificanceAttribute Mean No Yes Level

Equipment warranty 6.23 5.47 5.42 0.82Accuracy 6.08 5.93 6.32 0.14Technical support 5.94 5.87 6.09 0.32Reduction in labor costs 5.82 5.88 5.64 0.35Durability 5.73 5.67 5.87 0.39Simplicity of operation 5.69 5.69 5.64 0.84Tallying capabilities 5.52 5.33 5.89 0.03*Speed 5.46 5.26 5.79 0.03*Cost 5.28 5.38 5.02 0.16Sorting capabilities 5.19 5.00 5.68 0.01*Ability to switch species 4.89 4.84 4.89 0.89Compatibility 4.56 4.48 4.64 0.60Ability to switch grading rules 4.50 4.49 4.51 0.94Color sorting capabilities 4.35 4.24 4.53 0.39* Represents significant differences using analysis of variance

Have Scanning Tech Y/NMean Rating (1-7 scale)

Participants were asked to rate a list of methods of learning about new technology

as well. This question was also based on a (1-7) rating scale, where (1=least important,

7=most important). The methods provided included: tradeshows, plant visits, peer

discussions, trade journals, personal sales calls, advertisements, meetings and

67

symposiums, scientific journals, and unsolicited sales literature. Analysis of variance

(ANOVA) was used to test for significant differences between the two groups, past

adopters and non-adopters of hardwood scanning technology, at a confidence level of

95%.

Table 9 shows that some methods of learning about new technology rated below

the mean rating value of (4), while others were rated above (4). In addition, tradeshows

were rated highest overall at 5.26. Plant visits (5.19) and peer discussions (5.05) were

rated high as well. Advertisements, meetings and symposiums, scientific journals, and

unsolicited sales literature all rated below the mean rating value of (4). Significant

differences were detected between adopters and non-adopters of hardwood scanning

technologies on the personal sales calls, and scientific journal methods of learning about

new technologies equipment (Table 9). Significance levels of .02, personal sales calls,

and < .01, scientific journals were reported. Moreover, non-adopters rated both of these

methods of learning about new technologies higher than adopters, i.e., non-adopters rate

both personal sales calls and scientific journals higher in importance in learning about

new technologies than adopters.

68

Table 9: Mean ratings of methods of leraning about new technologies, adopters vs. non-adopters

Method of Learning Overall SignificanceAbout New Technology Mean No Yes Level

Tradeshows 5.26 5.27 5.29 0.93Plant visits 5.19 5.10 5.41 0.22Peer discussions 5.05 4.95 5.21 0.28Trade journals 4.42 4.50 4.19 0.19Personal sales calls 4.38 4.57 4.00 0.02*Advertisements 3.99 4.06 3.75 0.19Meetings, symposiums 3.76 3.77 3.74 0.92Scientific journals 3.23 3.43 2.75 <.01*Unsolicited sales literature 3.16 3.26 2.91 0.15* Represents significant differences using analysis of variance

Mean Rating (1-7 scale)Have Scanning Tech Y/N

CONCLUSIONS

Analysis showed dimension and flooring manufacturers view automated lumber

grading technology as more beneficial than small cabinet and furniture manufacturers.

Demographic information indicated that our cabinet and furniture samples represented a

subset of these populations; i.e., the cabinet and furniture samples represent smaller

companies in these two industries. In addition, in the case of dimension manufacturers,

the perceived benefit may be influenced by the significantly higher costs of grading

lumber they incur compared to cabinet and furniture manufacturers. In this scenario,

potential savings may be achieved through automated lumber grading technology. In the

case of flooring manufacturers, the perceived benefit may be influenced by the

significantly higher production volumes, which are characteristic of the flooring industry.

With increased lumber production volume more resources are required to grade lumber.

This results in higher lumber grading costs on a man-hour basis.

69

Companies currently using scanning technology in their production processes are

more likely to also market their products internationally than companies who do not have

scanning technology in place. International marketing activity is often considered a

characteristic of more innovative and progressive companies. Moreover, companies who

have adopted scanning technology are relatively larger companies both in terms of annual

sales and number of employees. The study found that the majority of companies who had

invested in scanning technology employed more than 100-200 people, and reported

greater than $5 million in annual sales. This finding was expected due to the capital

investment required to purchase scanning equipment. The study found that larger

companies with increased costs and resources are more likely to adopt scanning

technologies. It should be noted, however, that our cabinet and furniture samples were

predominantly comprised of smaller companies; therefore, we found these sectors to be

less likely to adopt scanning technologies, specifically automated lumber grading

technology.

Compared to other companies, those who have adopted scanning technologies

expect the cost of an automated lumber grading system to be higher. Moreover,

companies that had adopted scanning technologies found more importance in tallying

capability, sorting capability, and production speed. This may indicate areas where

further development or improvement of scanning equipment is necessary.

In addition, the dimension and flooring sectors tend to share information

concerning high-tech equipment more readily than the cabinets and furniture sectors.

The majority of respondents from the dimension and flooring sectors reported that they

70

had been contacted, concerning high-tech equipment, by other companies within their

industry. The opposite was found to be true with the cabinet and furniture sectors.

Therefore, information about the performance of this type of equipment will most likely

be shared between companies in the dimension and flooring sectors through word-of-

mouth.

In summary, potential adopters of automated lumber grading technology will

most likely be larger companies reporting greater than $5 million in annual sales and

employing more than 100 people. In addition, due to greater lumber throughput volumes

and greater costs associated with grading lumber, dimension and flooring manufacturers

may better represent potential adopters of automated lumber grading technology

compared to cabinet and furniture manufacturers; however, our sample is representative

of smaller cabinet and furniture manufacturers. Moreover, due to the degree to which

production is concentrated in the rough mill and the relatively high production volumes

associated with this industry, it is thought that large dimension manufacturers represent

the best market for automated lumber grading technology in the secondary wood products

industry. Overall, a market does exist for this technology; however, early adopters will

most likely be larger companies with sufficient resources to invest in high-tech

equipment.

71

REFERENCES Berkowitz, 1986. New Product Adoption by the Buying Organization: Who Are the Real Influencers? Industrial Marketing Management. Vol. 15, pages 33-43 Dempsey, Gilbert P. and W. G. Luppold. 1992. The State of Hardwood Lumber Markets. Northern Logger and Timber Processor. March, 3 pages. Hansen, Bruce and C. West. 1998. Trends in Domestic Hardwood Markets. Hardwood Symposium Proceedings. May 6-9, 9 pages. Hoff, Kristen, N. Fisher, S. Miller, and A. Webb. 1997. Sources of Competitiveness for Secondary Wood Products Firms: A Review of Literature and Research Issues. Forest Products Journal. Vol. 47, No. 2, pages 31-37. Kline, D. E., Connors R. W., and Araman P. A. 1998. What’s Ahead in Automated Lumber Grading, Proceedings of Scan-Pro Conference. Vancouver, British Columbia, Canada, 11 pages. Mullin, Sandy. 1998. Scanning Lumbr for Defects and Increased Profit. Wood & Wood Products. February, 1998, pages 37-38. Patterson, David W., Anderson, and Rockwell. 1997. Increased Use of Automated Machinery Requires Changes in Quality Control Procedures. Forest Products Journal. Vol. 47, No. 1, pages 33-36. Rosenburg, Nathan, P. Ince, K. Scog, and A. Plantinga. 1990. Understanding the Adoption of New Technology in the Forest Products Industry. Forest Products Journal. Vol. 40, No. 10, pages 15-22.

72

CHAPTER 3 ABSTRACT

In assessing markets for scanning technology in the secondary wood products industry it

is important to understand differences in machine attributes among different sectors of

the industry, and to identify the best methods of promotion to those sectors. For this

research scanning technology is defined as process optimization technology, which uses

various sensing technologies to perform such tasks as defect detection, optimized sawing,

and sorting and tallying of materials. Four industry sectors were surveyed in this study,

cabinets, furniture, dimension, and flooring, to identify the “benefit bundle” that would

increase the overall marketability of an automated lumber grading system (AGS). In this

study the cabinet and furniture samples were predominantly comprised of smaller

companies. In addition, for this study the benefit bundle is defined as the set of system

attributes and machine capabilities required by the secondary wood products industry.

This information is important to manufacturers of scanning technology in identifying

viable markets and determining the requirements of potential customers. Furthermore,

this research is important to the wood products industry in providing suppliers with

information specific to their businesses such that manufacturing innovations can move

more readily to the industry. The study found increased yield and efficiency to be

important system attributes. It was thought that automated lumber grading technology

could influence part yield by providing an optimum cut-up algorithm based on accurate

defect mapping and lumber grading. In addition, technical support and equipment

warranties were found to be important attributes to the secondary wood products

73

industry. Tradeshows were found to be important to the four sectors of the industry in

learning about new technology. Moreover, companies expressed interest in observing a

working system in an industrial setting before investing in this technology.

74

INTRODUCTION

The secondary wood products industry is a diverse industry manufacturing a wide

range of products, while utilizing different raw material mixes and production processes.

Products produced by this industry range from kitchen cabinets to railway ties.

Improving production efficiency and part yield are common interests among the many

companies in this industry. This study focused on scanning technology as a tool to

improve production efficiency and part yield, and to address other production concerns

such as improving the quality and consistency of products. For this research scanning

technology is defined as process optimization systems, which use various sensing

technologies to perform such tasks as defect detection, optimized sawing, and sorting and

tallying of materials. In identifying the benefit bundle that will increase the marketable

success of an automated lumber grading system, it is necessary to understand the

importance of system attributes among companies in the target market. For this research,

the benefit bundle is defined as the important system attributes and machine capabilities

required by companies in the target market. In this study, the target market was the

secondary wood products industry, specifically the wood cabinet, furniture, dimension,

and flooring industries. In identifying the most effective methods of promotion to these

sectors it is helpful to understand how companies in this industry receive information

concerning high-tech industrial products as well. This information is important in the

diffusion of scanning technology into the secondary wood products industry.

The secondary wood products industry is a large industry in terms of hardwood

lumber consumption. Greater than 90% of all hardwood lumber produced in the US is

75

consumed by eight secondary wood processing markets. These markets include pallets,

furniture, dimension parts, exports, millwork, flooring, cabinets, and railway ties.

Hardwood lumber consumption for these markets has been estimated at over 13 billion

board feet in 1997 (Hansen and West 1998). The four sectors in this study, cabinets,

furniture, dimension, and flooring, consumed approximately 52% of all hardwood lumber

produced in the US in 1997, an estimated 7.7 billion board feet. In addition, these sectors

showed an increase in hardwood lumber consumption in this decade (Hansen and West

1998). US consumption figures for 1997 show a total US hardwood lumber consumption

increase of over 2 billion board feet since 1991 (Dempsey and Luppold 1992).

From 1991 to 1997 hardwood lumber consumption in the cabinet industry

increased by 33% (Hansen and West 1998). According to Punches et al. (1995) the US

cabinet industry consists primarily of a relatively small number of large companies. In a

study by Adams (1999) it was noted that cabinet sales for use in rooms other than

kitchens and bathrooms has steadily increased throughout the 1990s. Eighteen of 36

cabinet companies surveyed responded that they sell other-room cabinets. In addition, it

was noted that the cabinet industry has experienced 33 consecutive months of growth in

sales since 1996, and both stock and custom manufacturers have averaged double-digit

growth in 1998.

The U.S. wood household furniture industry is a fragmented industry made up of

several large multi-plant firms and many small single-plant firms (West 1990). Recent

increases in imports have resulted in increased competition in the furniture industry.

“The U.S. is the largest net importer of furniture in the world” (Taylor 1999). In 1996

76

U.S. furniture imports showed a 71% increase over 1990 figures. This increase in

imports was valued at nearly US$7 billion (Taylor 1999). In addition, the wood furniture

industry showed a 50% increase in hardwood lumber consumption between 1991 and

1997 (Hansen and West 1998).

The dimension and wood components industry showed an 85% increase in

hardwood lumber consumption between 1991 and 1997 (Hansen and West 1998).

Dimension and wood components are defined as intermediary, value-added products,

which are used in other secondary wood processing operations such as millwork,

furniture, and cabinet manufacturing. “Wood components, such as dimension parts, are

dried and processed to a point where the maximum waste is left at the mill and the

maximum utility is delivered to the customer.” (Vlosky 1996). In some cases furniture

manufacturers own and operate their own dimension mills, and often dimension mills are

affiliated with sawmills, but most are independent companies. According to Vlosky,

(1996) dimension and component manufacturers are often the focus of state economic

development initiatives in an attempt to add value to hardwood resources.

From 1991 to 1997 hardwood lumber consumption in the flooring industry

increased by 120% (Hansen and West 1998). Recently wall-to-wall carpet and PVC

flooring have begun to lose their market share to wood flooring as a result of increasing

health and environmental concerns among consumers (Weinig 1999). Weinig states,

“Reservations about chemicals and fear of house dust by many allergy sufferers has led

to a decrease in carpet sales versus wood flooring.” In addition, consumer preferences

77

have shifted more toward wood flooring products in relation to home interior design

fashion in recent years.

The four industry sectors previously outlined represent potential markets for

scanning technology. These sectors produce different products, and utilize different raw

material mixes and production processes. In assessing markets for automated lumber

grading technology, differences in system requirements between industry sectors and

effective promotion methods must be identified. From the marketing perspective, the

adoption of industrial high technology products can be evaluated on their perceived

attributes. Mittal et al. (1998) state that negative performance on an attribute has a

greater impact on overall satisfaction than positive performance on the same attribute.

Overall customer satisfaction with a product shows diminishing sensitivity to product

attribute-level performance; i.e., as overall satisfaction increases, customers become less

sensitive to specific product attributes or more tolerant to deficiencies related to specific

attributes. By installing systems that possess attributes desirable to specific

manufacturing operations, overall satisfaction level will increase.

Differences in system attributes and machine requirements were expected to occur

between the industry sectors in this study due to the variation in manufacturing processes,

products, raw materials, lumber grades, and company demographics that characterize

each sector. For example, the wood furniture industry manufactures a wider range of

products, which typically require more intensive processing compared to the wood

flooring industry. In addition, the wood furniture industry utilizes a lumber grade mix

consisting predominantly of lumber graded No. 1 Common and higher (Hansen et al.

78

1995). This scenario is different from the flooring industry, which utilizes a mix of lower

grades of lumber, No. 1 Common and below (Hansen and West 1998). It was

hypothesized that differences such as these would contribute to variation in technology

requirements among the four groups in this study.

OBJECTIVES

The objectives of this study were:

1. to characterize the benefit bundle that will increase the overall marketable success

of and automated lumber grading system, and

2. to identify the best methods of promoting this technology to the secondary wood

products industry.

METHODOLOGY

A questionnaire was developed based on the study objectives. This questionnaire

was reviewed by faculty of Virginia Polytechnic Institute and State University for

relevance to the objectives. The questionnaire was further developed in personal

interviews at one company in each industry sector. These interviews were used to

identify important information that should be included in the questionnaire. Further

adjustments were made to the questionnaire and it was pre-tested by fax with three

companies in each industry sector. The questionnaire was pre-tested to ensure that

questions were clear and followed a logical format. Participants were asked to return the

79

completed survey along with any suggested changes. Final adjustments were made

following the pretest, and the revised questionnaire was mailed in May, 1999.

A total of 1,256 identical questionnaires were mailed out to the sample frame

consisting of four sectors of the secondary wood products industry; cabinets, furniture,

dimension, and flooring manufacturers. The sample is predominantly comprised of

members of various trade associations. It was assumed that members of trade associations

would represent the more innovative companies in the industry; therefore, by sampling

from this group there could have been bias introduced toward more innovative

companies. In addition, this study assumes that relatively innovative firms represent the

most likely adopters of an automated lumber grading system, hence the sample bias is

justified. A sample consisting of companies not likely to adopt this technology would

conflict with our stated objectives and potentially negatively influence our results and

conclusions.

RESULTS

Survey data was entered into SPSS statistical software package for analysis when

received. A total of 439 responses were received. Of those, 209 questionnaires were

useable, resulting in an overall response rate of 16.64%. Adjusting the overall response

rate for undeliverable questionnaires (175), and companies performing operations not

applicable to the sample (55), the overall adjusted response rate was calculated at 20.4%.

The numbers of survey questionnaires received according to business type were, cabinets

(45), furniture (35), dimension (92), and flooring (37). These responses resulted in

80

response rates of 11.6% for cabinet manufacturers, 9% for furniture manufacturers,

23.7% for dimension manufacturers, and 40.2% for flooring manufacturers. Adjusting

the response rates per industry segment resulted in 13.6% for cabinet manufacturers,

12.4% for furniture manufacturers, 28.4% for dimension manufacturers, and 45.1% for

flooring manufacturers.

A sample size of 97 respondents was calculated based on a 25% response rate for

the cabinets, furniture, and dimension sectors. The flooring sector consists of a known

population that is smaller than the calculated sample size of 97; therefore, the entire

known population of US flooring manufacturers was surveyed.

Moreover, due to the relative frequency with which these industry segments are

surveyed we did not receive the anticipated 97 responses per industry sector. Smith

(1999) states that it is necessary to have more observations per group than total variables

for adequate statistical power in MANOVA. Our largest number of variables for a rated

question was 14 well below our smallest number of observations for one group, 35. In

addition, our total number of observations was adequate for MANOVA at 209. The

MANOVA was repeated at a confidence level of 90%, which generated results that were

identical to those from the 95% test. Significance levels were similar for the two tests;

therefore, the conclusions formulated in this study are statistically valid.

THE BENEFIT BUNDLE

The first objective of this study was to identify the benefit bundle that would

increase the marketable success of an automated lumber grading system in the secondary

wood products industry. Differences in the importance of various machine requirements

81

were expected to exist between industry sectors due to inherent variability associated

with producing different products. A list of factors dealing with system attributes and

machine capabilities was developed, and respondents were asked to rate the importance

of each factor in the list in their decision to adopt an automated lumber grading system.

Factors included in the questionnaire were accuracy, cost, durability, compatibility with

existing equipment, throughput speed, ability to switch grading rules, ability to switch

species, simplicity of operation, technical support, equipment warranty, reduction in

labor costs, tallying capabilities, sorting capabilities, and color sorting capabilities. The

rating hierarchy for each factor was based on a (1-7) scale where (1=least important,

7=most important). Multivariate analysis of variance (MANOVA) was used to detect

statistically significant differences between industry sectors on the 14 factor ratings.

Fisher’s LSD technique was used as a post hoc test to detect significant differences

between specific groups. MANOVA analysis was used to detect significant differences

among industry sectors at a confidence level of 95%. A confidence level of 95%

indicates an alpha level, or probability of type 1 error, equal to .05. This means that we

are 95% confident that the actual mean rating values of the population studied, on

average will be equal to the mean rating values reported in the statistical tests.

All sectors rated every factor in the list at or above the average importance value

of (4) indicating that all 14 factors were at least of average importance when considering

the adoption of an automated lumber grading system. Of the 14 attributes listed,

equipment warranty rated highest, on average, among all respondents. A mean rating

value of 6.23, on a (1-7) rating scale, was reported for this attribute among all

82

respondents in the study. Accuracy was rated second highest among all respondents with

a mean rating value of 6.08. Among the lowest rated factors were switching species

(4.89), compatibility (4.56), switching grading rules (4.50), and color sorting capabilities

(4.35) (Table 10).

Table 10 shows significant differences detected between groups on the equipment

warranty, accuracy, technical support, reduction in labor costs, durability, simplicity of

operation, tallying capabilities, throughput speed, machine cost, sorting capabilities,

ability to switch species, compatibility with existing equipment, ability to switch grading

rules, and color sorting capabilities factors. Multivariate significance levels calculated

for these factors were; less than .01 for the accuracy factor, .04 for the durability factor,

.06 for the compatibility factor, and .02 for the speed factor. Other significance levels

were species switching ability (.04), reduction in labor costs (.04), tallying capabilities (<

.01), and sorting capabilities (.03).

When running multiple comparisons on each sector and factor, a significant

difference was detected between the furniture and cabinet sectors on the accuracy factor

(Table 10). A significance level of .05 was reported for this comparison. This indicates

that these industry sectors differ in their rating of the importance of accuracy in the

decision to adopt an automated lumber grading system. In addition, the cabinet sector

reported a higher mean rating value of 5.74 compared to 4.84 for furniture; therefore the

cabinet sector places a significantly higher degree of importance on this factor than the

furniture sector. Differences were also detected between the furniture and dimension

sectors and the furniture and flooring sectors on this factor. Strongly significant values

83

for both comparisons were calculated at less than .01. A strongly significant difference

such as this indicates a strong difference in the importance rating of these sectors on the

accuracy factor. In both comparisons dimension and flooring manufacturers reported

substantially higher mean rating values than furniture manufacturers for this factor, 6.35

for dimension and 6.46 for flooring compared to 4.84 for furniture.

Significant differences were detected between the cabinet and dimension, and

furniture and dimension sectors on the reduction in labor costs factor (Table 10).

Significance levels for these comparisons were calculated at .03 for the cabinet sector

compared to the dimension sector and .04 for the furniture sector compared to the

dimension sector. In both cases dimension manufacturers reported a higher mean rating

value, 6.13 compared to 5.37 for cabinet manufacturers and 5.32 for furniture

manufacturers, respectively. This indicates that dimension manufacturers consider this

factor to be more important than cabinet and furniture manufacturers in the decision to

adopt automated lumber grading technology. These findings are consistent with the high

lumber grading costs associated with the dimension and flooring industries.

Significant differences were detected between cabinet and flooring, and furniture

and flooring manufacturers on the machine durability factor (Table 10). Significance

levels were calculated at .03 and .01 for the two comparisons respectively. Moreover, the

flooring sector reported a higher mean value in both comparisons, 6.11 for flooring

manufacturers compared to 5.37 for cabinet manufacturers and 5.16 for furniture

manufacturers. Therefore, on average, flooring manufacturers rate machine durability

higher than cabinet and furniture manufacturers in their decision to adopt automated

84

lumber grading technology. These findings are consistent with demographic information

showing the relatively high throughput of lumber associated with the dimension and

flooring industries.

In multiple comparisons between sectors on the tallying capabilities factor,

significant differences were detected between the cabinet and dimension, and cabinet and

flooring sectors, and furniture and dimension, and furniture and flooring sectors (Table

10). Significance levels for these comparisons were calculated at .02 for the cabinet

sector compared to the dimension sector, .01 for cabinet compared to flooring, < .01 for

furniture compared to dimension, and < .01 for the furniture sector compared to the

flooring sector. The dimension and flooring sectors reported higher mean rating values,

5.8 for dimension and 5.97 for flooring compared to 5.0 for cabinet and 4.16 for

furniture. Moreover, strong significance levels calculated for the furniture and

dimension, and furniture and flooring comparisons indicate a strongly significant

difference in the importance ratings of these sectors on this factor.

Significant differences were detected between the furniture and dimension,

cabinet and flooring, and furniture and flooring sectors on the production speed factor

(Table 10). Significance levels for these comparisons were calculated at .02 for the

furniture sector compared to the dimension sector, .04 for the cabinet sector compared to

the flooring sector, and < .01 for the furniture sector compared to the flooring sector. A

significance level of < .01 indicates a strongly significant difference. The dimension

sector reported a higher mean rating value than the furniture sector, 5.59 compared to

4.74, indicating that dimension manufacturers rate this factor higher than furniture

85

manufacturers. Moreover, the flooring sector reported a higher mean rating value than

the cabinet and furniture sectors, 5.89 compared to 5.11 for the cabinet sector and 4.74

for the furniture sector. In the case of the furniture and flooring comparison, the strongly

significant difference indicates a strong difference in the importance of speed in the

decision to adopt automated lumber grading technology. As with the durability factor,

these findings are consistent with the high volumes of lumber throughput associated with

the dimension and flooring industries.

Significant differences were detected between the furniture and cabinet, furniture

and dimension, and furniture and flooring sectors on the sorting capabilities factor (Table

10). Significance levels for these comparisons were calculated at .03 for the furniture

sector compared to the cabinets sector, < .01 for the furniture sector compared to the

dimension sector, and .03 for the furniture sector compared to the flooring sector. In all

comparisons the furniture sector reported a lower mean rating value than the other

industry sectors, 4.16 compared to 5.22 for cabinet, 5.38 for dimension, and 5.14 for

flooring. This indicates that furniture manufacturers place a lower degree of importance

on this factor than cabinet, dimension, and flooring manufacturers. Moreover, the strong

significance level calculated for the furniture and dimension comparison indicates a

stronger degree of difference in the importance placed on sorting capabilities by the

dimension sector compared to the furniture sector.

86

Table 10: Mean ratings of machine attributes according to industry sector System Overall Significance

Attribute Mean Cabinets (1) Furniture (2) Dimension (3) Flooring (4) LevelEquipment warranty 6.23 4.74 6.68 6.85 5.83 0.46Accuracy 6.08 5.74 4.84 1 3 4 6.35 6.46 <.01*Technical support 5.94 5.52 5.42 6.13 6.11 0.07Reduction in labor costs 5.82 5.37 5.32 6.13 1 2 5.57 0.04*Durability 5.73 5.37 5.16 5.81 6.11 1 2 0.04*Simplicity of operation 5.69 5.59 5.42 5.68 5.89 0.69Tallying capabilities 5.52 5.00 3 4 4.16 3 4 5.80 5.97 <.01*Speed 5.46 5.11 4.74 3 5.59 5.89 1 2 0.02*Cost 5.28 5.04 5.32 5.39 5.20 0.76Sorting capabilities 5.19 5.22 4.16 1 3 4 5.38 5.14 0.03*Ability to switch species 4.89 4.85 4.05 5.29 2 4 4.46 0.04*Compatibility 4.56 4.59 3.53 1 3 4 4.65 4.91 0.05*Ability to switch grading rules 4.50 4.96 4.00 4.62 4.09 0.17Color sorting capabilities 4.35 5.04 3.95 4.30 4.11 0.231,2,3,4 Represent significant differences between specific groups* Represents MANOVA significance, at least one difference among groups

Mean Rating (1 - 7 scale)

In multiple comparisons between sectors on the ability to switch species factor,

significant differences were found between the furniture and dimension, and flooring and

dimension sectors (Table 10). Significance levels for these comparisons were calculated

at .01 for the furniture sector compared to the dimension sector and .04 for the flooring

sector compared to the dimension sector. In both cases dimension manufacturers

reported a higher mean rating, 5.29 compared to 4.05 for the furniture sector and 4.46 for

the flooring sector. This indicates that dimension manufacturers rate the importance of

this factor higher than furniture and flooring manufacturers in the decision to adopt

automated lumber grading technology.

In multiple comparisons between sectors on the compatibility factor, significant

differences were detected between the furniture and cabinet, furniture and dimension, and

furniture and flooring sectors (Table 10). Significance levels for these comparisons were

calculated at .05, for the furniture and cabinet comparison, .02, for the furniture and

dimension comparison, and < .01, for the furniture and flooring comparison. The

87

furniture sector reported a lower mean rating value in all comparisons, 3.53 for the

furniture sector compared to 4.59 for the cabinet sector, 4.65 for the dimension sector,

and 4.91 for the flooring sector. This indicates that the cabinet, dimension, and flooring

sectors all consider the compatibility factor to be more important than the furniture sector.

Most significant differences found to exist between industry sectors were

differences between the cabinet and furniture sectors, and dimension and flooring sectors

in general. Overall, the dimension and flooring sectors reported higher mean rating

values on all attributes for which a difference was found. These findings could be

translated into an overall lack of interest or need in this technology by the furniture and

cabinet industries, or an overall higher interest or need by the dimension and flooring

industries. It should be noted that our cabinet and particularly our furniture samples were

comprised of predominantly smaller companies; therefore, conclusions will be relevant to

smaller companies in the cabinet and furniture industries.

Moreover, while equipment warranty, accuracy, technical support, reduction in

labor costs, and machine durability rated high overall among the four sectors, cabinet,

furniture, dimension, and flooring, some differences were present among the optimal

benefit bundle characterized for each sector. Table 11 illustrates that the cabinet sector

rated accuracy (5.74), simplicity of operation (5.59), technical support (5.52), reduction

in labor costs (5.37), and machine durability (5.37) among the fourteen variables

provided. The furniture sector reported mean ratings of 6.68 for equipment warranty,

5.42 for technical support, 5.42 for simplicity of operation, 5.32 for reduction in labor

costs, and 5.32 for machine costs. Top mean ratings for the dimension sector included,

88

equipment warranty (6.85), accuracy (6.35), technical support (6.13), reduction in labor

costs (6.13), and machine durability (5.81). The flooring sector reported top mean

ratings of 6.46 for accuracy, 6.11 for technical support, 6.11 for machine durability, 5.97

for tallying capabilities, and throughput speed and simplicity of operation were both

rated at 5.89 among the fourteen factors provided. In addition, the flooring sector was the

only one to cite production speed and tallying capability as attributes in the bundle. The

furniture sector was the only one to cite machine cost as an important attribute.

Table 11: Optimum benefit bundle according to industry sector

M e a n R a tin g(1 -7 s c a le )

S y s te m A ttr ib u te C a b in e tsA c c u ra c y 5 .7 4S im p lic ity o f o p e ra tio n 5 .5 9T e c h n ic a l s u p p o rt 5 .5 2R e d u c tio n in la b o r c o s ts 5 .3 7M a c h in e d u ra b ility 5 .3 7

F u rn itu reE q u ip m e n t w a rra n ty 6 .6 8T e c h n ic a l s u p p o rt 5 .4 2S im p lic ity o f o p e ra tio n 5 .4 2R e d u c tio n in la b o r c o s ts 5 .3 2M a c h in e c o s t 5 .3 2

D im e n s io nE q u ip m e n t w a rra n ty 6 .8 5M a c h in e a c c u ra c y 6 .3 5T e c h n ic a l s u p p o rt 6 .1 3R e d u c tio n in la b o r c o s ts 6 .1 3M a c h in e d u ra b ility 5 .8 1

F lo o rin gM a c h in e a c c u ra c y 6 .4 6T e c h n ic a l s u p p o rt 6 .1 1M a c h in e d u ra b ility 6 .1 1T a lly in g c a p a b ilitie s 5 .9 7T h ro u g h p u t s p e e d 5 .8 9S im p lic ity o f o p e ra tio n 5 .8 9

89

Participants were asked to rate a list of 3 factors (invoice verification, better

matching of grade to product, and facilitation of lumber sorting) on their importance to

lumber grading in their companies. A (1-7) rating scale was used where (1=least

important and 7=most important). Mean rating values for all sectors were greater than

the average importance rating of (4) indicating that all four sectors in the study found the

listed factors to be of greater than average importance in lumber grading (Table 12). In

addition, the highest mean rating value was reported for the invoice verification factor at

5.45.

Multivariate analysis of variance (MANOVA) was used to make comparisons

between sectors on the listed factors. The MANOVA was used to test for significant

differences between industry sectors at a confidence level of 95%. MANOVA resulted in

no significant differences between the four sectors in this study on the three listed factors.

Significance levels showed no difference among the industry sectors on the three

variables. This indicates that the four sectors did not significantly differ in their

importance rating of the listed factors as they relate to lumber grading. While there were

no differences in the reasons for grading lumber among the four sectors, invoice

verification was rated most important among the listed factors. Verification of lumber

invoices is important to companies to ensure that grade mixes and payments to vendors

are consistent with orders.

Table 12: Mean ratings of factors in lumber grading according to industry sector Overall SignificanceMean Cabinets (1) Furniture (2) Dimension (3) Flooring (4) Level *

Invoice verification 5.45 5.04 5.35 5.45 5.94 0.24Matching grade to product 5.33 5.87 4.88 5.51 5.06 0.15Facilitates lumber sorting 4.37 4.29 4.35 4.41 4.44 0.99* Multivariate analysis of variance

Mean Rating (1 - 7 scale)Influential Factors

90

In addition to determining the reasons for grading lumber among the industry

sectors, participants were asked to indicate those factors that influenced their decision to

invest in scanning technology in the past, if applicable. Included in the list were

increased efficiency, automation, yield, and accuracy. Table 13 shows that 100% of

cabinet manufacturers who responded noted that increased efficiency was a factor in their

decision to invest in scanning technology. Eighty percent of the furniture industry,

75.7% of the dimension industry, and 61.5% of the flooring industry noted efficiency as a

factor influencing their decision to invest in scanning technology. Therefore, the ability

to increase production efficiency is an important machine capability to the secondary

wood products industry.

Greater than 71% of cabinet manufacturers and 80% of the furniture sector noted

automation as a factor in their decision to invest in scanning technology, while in contrast

only 48.6% of dimension manufacturers and 38.5% of flooring manufacturers noted

automation as a factor. These results could have been influenced by the relative smaller

size of most companies in the cabinet and furniture samples; i.e., the desire to increase

production automation may be perceived as a higher priority among smaller

manufacturers. As expected, increased yield influenced the majority of companies in all

four sectors in their decision to invest in scanning technology in the past. Greater than

71% of the cabinet sector, 80% of the furniture sector, 89.2% of the dimension sector,

and 92.3% of the flooring sector responded that yield was an influential factor in their

decision to invest in scanning technology in the past. It is thought that automated lumber

grading technology could influence yield by providing an optimum cut-up algorithm

91

through accurate grading and defect mapping. These findings indicate that while

increased automation is more important to smaller cabinet and furniture sectors, the

potential to increase yield is an important factor to all industry sectors.

In addition to identifying the factors influencing adoption of scanning technology,

it is important to understand where in the operation such equipment would be most

beneficial. This discussion focuses on the location or locations inside the plant where an

automated lumber grading system should be installed. It was thought that an automated

lumber grading system would logically fit best where lumber was being graded. To

determine the location inside the plant where an automated lumber grading system would

be most beneficial, participants were asked to indicate at what point or points in their

operation lumber was graded. Participants were asked to indicate various locations from

a list provided. The list included at delivery, first-in-first-out, before drying, and after

drying. Table 13 illustrates the majority of all four sectors indicated that they graded

lumber at least at delivery. Greater than 72.7% of cabinet manufacturers, 63.3% of

furniture manufacturers, 55.7% of dimension manufacturers, and 68.8% of flooring

manufacturers graded lumber upon delivery to the plant. Among the four sectors,

dimension manufacturers graded the largest amount of lumber before drying, 26.1%.

This factor was of course influenced by the presence of dry kilns at the plant.

It was found that many companies graded lumber more than once in their

operation. The cabinet sector was found to also grade lumber on a first-in-first-out basis

and in some cases after drying the lumber. Some companies in the furniture and flooring

sectors graded lumber at three points in the production process, first-in-first-out, before

92

drying, and after drying. The dimension sector reported some companies grading lumber

before and after drying in addition to grading at delivery.

Table 13: Percentage of adoption factors, location of lumber grading, and expected machine cost according to industry sector

Influential Factor inAdoption of Scanning Tech Cabinets Furniture Dimension Flooring Overall

Increased efficiency 100% 80% 75.7% 61.5% 75.8%Automation 71.4% 80% 48.6% 38.5% 51.6%Yield 71.4% 80% 89.2% 92.3% 87.1%Accuracy 71.4% 40% 43.2% 41.7% 45.9%

Points of Lumber Grading OverallAt delivery 72.7% 63.3% 55.7% 68.8% 62.3%First-in-first-out 15.2% 13.3% 5.7% 9.4% 9.3%Before drying 3% 16.7% 26.1% 18.8% 19.1%After drying 9.1% 6.7% 12.5% 3.1% 9.3%

Expected Machine Cost OverallLess than $50,000 38.5% 33.3% 17.7% 6.1% 20.8%$50,000 - $150,000 34.6% 38.1% 39.2% 39.4% 38.4%$150,001 - $250,000 11.5% 14.3% 24.1% 15.2% 18.9%$250,001 - $500,000 11.5% 14.3% 12.7% 24.2% 15.1%Greater than $500,000 3.8% 0 6.3% 15.2% 6.9%

Response Frequency

Once the factors influencing adoption and the most beneficial location in the plant

were identified, it was considered important to determine the perceived value of

automated lumber grading technology. In determining the expected cost of an automated

lumber grading system participants were asked to indicate, from a list of price ranges,

what they would expect an automated lumber grading system, with all the desired

features, to cost. A range of less than $50,000 to greater than $500,000 was provided.

The majority of companies in the four sectors, except for the flooring sector, responded

with expected machine costs of $150,000 or less (Table 13). Greater than 73% of cabinet

manufacturers, 71.4% of furniture manufacturers, 56.9% of dimension manufacturers,

and 45.5% of flooring manufacturers expected an automated lumber grading system to

cost less than $150,000. Moreover, the furniture sector was the only sector in the study

in which no companies responded in the greater than $500,000 range. The flooring sector

93

reported the largest percentage of companies, 15.2%, expecting an automated lumber

grading system to cost in excess of $500,000.

We asked participants to list any factors that would prevent them from purchasing

an automated lumber grading system, assuming that the technology was proven and

effective. This was an open-ended question. A total of 61 responses were received,

29.2% of total responses. Of those who responded, 42.6% reported cost as a factor that

would prevent them from purchasing an automated lumber grading system. Other factors

listed were inability to meet target payback, lack of need, lack of plant space, and lack of

sufficient technical support.

We then asked participants to list factors that would convince them to invest in an

automated lumber grading system. A total of 75 responses were received, 35.9% of total

responses. Of those who responded, 16.7% reported affordable cost as a factor that

would influence them to invest in an automated lumber grading system. Approximately

10% listed meeting their target payback, and 14.2% reported that observing a proven

system in use in an industrial setting would convince them to purchase an automated

lumber grading system.

Participants were asked to cite specific reasons why they were or were not

satisfied with technology equipment in their plants, if applicable. A total of 170

responses were received, 81.3% of total responses. There were no prevailing responses

reported. Responses were comprised of various factors. Labor savings, increases in

production speeds, increased production efficiency, and competitive advantage were

listed among those who were satisfied. Responses from those who were not satisfied

94

included poor or non-existent technical support, system not yet refined, poor reliability,

system was marketed before fully developed, low durability, and information not

delivered in usable form. These findings are consistent with the high ratings of technical

support and machine warranty among participants. This information provides further

justification as to the importance of service related factors as they relate to automated

lumber grading technology. In addition, these findings suggest that perhaps these types

of systems had not been fully developed and tested before being marketed in the past.

In summary, significant differences were found to exist between the cabinet and

furniture, and dimension and flooring sectors, in general, on machine durability, machine

speed, reduction in labor costs, and tallying capabilities. The dimension and flooring

sectors rate these attributes higher than the cabinet and furniture sectors. These findings

are consistent with the greater volumes of lumber throughput and increased costs

associated with the dimension and flooring sectors compared to the cabinet and furniture

sectors. Again, it should be noted that the cabinet and furniture samples were comprised

of mostly smaller manufacturers. In addition, increased part yield was rated high among

all industry sectors. It was also found that the four sectors in this study do not differ

significantly in their reasons for grading lumber. Invoice verification was the most

important factor in lumber grading among all sectors. Moreover, system cost, technical

support, and the ability to observe a system in an industrial environment were cited as

important factors in the decision to invest in automated lumber grading technology. In

addition, an automated lumber grading system must be able to meet short payback

periods, and be proven to reduce overall production costs.

95

PROMOTION

The second objective of this research was to determine the best methods of

promoting scanning technology, specifically an automated lumber grading system, to the

secondary wood products industry. Differences in the ways in which different industry

sectors receive information about technology equipment were expected to exist. It was

hypothesized that the four sectors in this study would differ in the degree to which they

shared technology information, and in the sources from which they sought information.

A list of nine methods of learning about new technology was developed and respondents

were asked to rate each method in importance in learning about new technology. The

rating was based on a 1-7 scale (1=least important, 7=most important). Methods included

in the questionnaire were: advertisements, trade shows, plant visits, personal sales calls

from manufacturers, meetings and symposiums, trade journals, scientific journals, peer

discussions, and unsolicited sales literature.

Multivariate analysis of variance (MANOVA) was used to make comparisons

among the four industry sectors in this study on the nine variables listed. Fisher’s LSD

technique was used as a post hoc test to detect significant differences between specific

sectors. The MANOVA analysis was used to test for significant differences between

industry sectors at a confidence level of 95%.

Overall mean rating values for all nine methods of learning about new

technology show trade shows, on average, rating the highest among all respondents at

5.26 (Table 14). In addition, plant visits and peer discussions were among the top three

highest rated methods at 5.19 and 5.05. Meetings and symposiums (3.76), scientific

96

journals (3.23), and unsolicited sales literature (3.16) all rated below average in learning

about new technology, among respondents, i.e., mean rating values for these methods

were below the mean rating value of (4), indicating that these methods were not

important in learning about new technology.

Significant differences were detected between sectors on the trade shows and

advertisement methods of learning about new technology. Multivariate significance

levels were calculated at .03 for trade shows and < .01 for advertisements. In addition,

most methods overall were rated higher than the mean rating value of (4).

When running multiple comparisons on each sector and method, a significant

difference was reported between the furniture and flooring sectors, furniture and

dimension sectors, and cabinet and dimension sectors on the advertisement method of

learning about new technology. In comparing the furniture and flooring sectors on

advertisements, a significance level of .03 was calculated. In addition, the furniture

sector reported a higher mean rating value (4.85 compared to 4.03) for this method. This

indicates that, on average, furniture manufacturers rely more on advertisements when

learning about new technology than flooring manufacturers.

In addition, a significant difference was detected between the furniture and

dimension sectors as well. A significance level of < .01 was calculated for this

comparison. Again, the furniture sector reported a higher mean (4.85 compared to 3.54).

This indicates that a strong difference exists in the degree of importance placed on

advertisements in learning about new technology between furniture and dimension

manufacturers. Furniture manufacturers rely significantly more heavily on this method

97

than dimension manufacturers. Finally, a significant difference was detected between the

cabinet and dimension sectors on the advertisement method as well. A significance level

of .02 was calculated for this comparison. The cabinet sector reported a higher mean

rating value (4.28 compared to 3.59); therefore it can be stated that cabinet manufacturers

place more importance on advertisements than dimension manufacturers in learning

about new technology.

In multiple comparisons between sectors on the trade show method of learning

about new technology, significant differences were detected between the furniture and

dimension, and furniture and flooring sectors. Significant levels for these comparisons

were .02 for the furniture sector compared the dimension sector, and < .01 for the

furniture sector compared to the flooring sector. In both cases the furniture sector

reported a higher mean rating value (5.96) indicating that furniture manufacturers rely

more heavily on trade shows in learning about new technology than dimension (5.16) and

flooring (4.76) manufacturers. In the case of furniture manufacturers compared to

flooring manufacturers, the difference between the two sectors was strongly significant.

Table 14: Mean ratings of methods of learning about new technology according to sector Overall SignificanceMean Cabinets (1) Furniture (2) Dimension (3) Flooring (4) Level

Trade shows 5.26 5.22 5.96 3 4 5.16 4.76 0.03*Plant visits 5.19 4.78 5.00 5.32 5.21 0.40Peer discussions 5.05 5.00 5.15 5.15 4.79 0.67Trade journals 4.42 4.47 4.85 4.32 4.09 0.28Personal sales calls 4.38 4.47 4.08 4.28 4.59 0.57Advertisements 3.99 4.28 4.85 4 3.59 1 2 4.03 <.01*Meetings, sypmposiums 3.76 3.44 4.15 3.79 3.56 0.34Scientific journals 3.23 3.19 3.23 3.24 3.18 1.00Unsolicited sales literature 3.16 3.16 3.31 3.16 3.03 0.921,2,3,4 Represent significant differences between specific groups* Represents MANOVA significance, at least one difference among groups

Mean Rating (1 - 7 scale)Method Learning New Tech

98

Analysis of overall response indicates that not only were trade shows, plant visits,

and peer discussions of overall more importance in learning about new technology, but

these three methods were rated the top three methods for each individual industry sector

as well. Table 15 shows cabinet manufacturers rated trade shows highest at 5.22, peer

discussions at 5.00, and plant visits at 4.78. Furniture manufacturers rated trade shows

highest as well at 5.96, while the peer discussion method was rated at 5.15, and plant

visits were rated at 5.00 among the nine methods provided. The dimension sector rated

plant visits highest at 5.32, while trade shows were rated at 5.16, and peer discussions

were rated at 5.14. The flooring sector rated plant visits highest as well at 5.21, while

peer discussions were rated at 4.79, and trade shows were rated at 4.76 among the nine

methods provided.

Table 15: Important methods of learning about new technology according to industry sector

Mean RatingMethod of learning about (1-7 scale)

New Technology CabinetsTradeshows 5.22Peer discussions 5.00Plant visits 4.78

FurnitureTradeshows 5.96Peer discussions 5.15Plant visits 5.00

DimensionPlant visits 5.32Tradeshows 5.16Peer discussions 5.15

FlooringPlant visits 5.21Peer discussions 4.79Tradeshows 4.76

99

The furniture and cabinet sectors reported nearly equivalent percentages of yes

and no answers to the following question: Do companies within your industry ever

contact you to inquire about high-tech equipment you have or have not installed? Greater

than 52% of the cabinet sector responded that they had not been contacted concerning

high-tech equipment they were considering, while 55.6% of the furniture sector reported

that they had not been contacted. In contrast, the dimension and flooring sectors reported

that 60.2% and 54.1% of companies respectively had been contacted concerning high-

tech equipment. This indicates that information is shared more freely among competitors

in the dimension and flooring sectors compared to the cabinets and furniture sectors.

This may have an influence on the appropriate method of promotion to different sectors

of the industry.

CONCLUSIONS

This study sought to identify the benefit bundle that would increase the

marketable success of automated lumber grading technology, and determine the most

effective methods of promoting scanning technology, specifically automated lumber

grading technology, to the secondary wood products industry. These four sectors of the

secondary wood products industry, small cabinet and furniture manufacturers, dimension,

and flooring manufacturers, expect to pay no more than $150,000 to $250,000 for an

automated lumber grading system. For most companies in these sectors of the industry,

high equipment costs would prevent them from purchasing an automated lumber grading

system. Furthermore, the ability to meet target payback periods, and the ability to see a

100

proven system in use in an industrial setting would be important factors in their adoption

of this technology.

In addition, equipment warranty and technical support rated among the top three

attributes in overall importance among all respondents. This finding indicates that

industry concerns, when considering this type of technology are not only equipment

related, but include services provided by the manufacturer as well. The ability to increase

part yield is an important concern in all four sectors; therefore, this function would be an

important component of the benefit bundle. It is thought that automated lumber grading

technology could influence yield by providing an optimum cut-up algorithm through

accurate grading and defect mapping thereby optimizing part yield from lumber.

Machine durability, throughput speed, reduction in labor costs, and tallying capabilities

are more important to dimension and flooring manufacturers than smaller cabinet and

furniture manufacturers in the decision to adopt automated lumber grading technology.

This finding is consistent with the increased costs, both monetarily and in terms of time,

which characterize the dimension and flooring industries.

Cabinet manufactures indicated that the optimum benefit bundle for their needs

would include high system accuracy, technical support, simple operation, the ability to

reduce labor costs, and high durability. Furniture manufactures responded with system

warranty, technical support, simple operation, ability to reduce labor costs and

affordable machine cost as important attributes. Dimension manufacturers indicated that

the optimum benefit bundle for their needs would include a system warranty, high system

accuracy, technical support, ability to reduce labor costs, and high machine durability.

101

Finally, flooring manufacturers responded with high system accuracy, technical support,

high machine durability, tallying capability high production speed, and simple operation

as attributes in the optimum benefit bundle.

The furniture industry was found to place more importance on advertisements

and trade shows than other sectors of the industry. It should be noted that our cabinet and

furniture samples consisted of smaller companies; therefore, conclusions that relate to

these two sectors represent small companies only. Trade shows are of overall importance

along with plant visits and peer discussions in learning about new technology among the

secondary wood products industry. In addition, the four sectors individually rated these

three methods highest in learning about new technology; therefore, these methods should

be utilized in promoting to the cabinet, furniture, dimension, and flooring sectors.

Advertisements, meetings and symposiums, scientific journals, and unsolicited sales

literature are of little importance in learning about new technology among the cabinet,

furniture, dimension and flooring sectors; therefore, these methods would not be effective

in promotion to these four sectors of the industry.

This study found that more dimension and flooring manufacturers, on average,

receive contact by other companies concerning high-tech equipment than smaller cabinet

and furniture manufacturers. This may be representative of the degree of perceived

innovativeness among companies in the dimension and flooring industries. In addition,

this factor suggests more open lines of communication between companies in these

sectors of the industry. It would be beneficial for manufacturers of scanning technologies

to form partnerships with members of the secondary wood products industry in

102

developing and testing this technology. Moreover, based on responses from past adopters

of scanning technologies, important system attributes to the focus of promotional efforts

include the ability of the system to increase production efficiency and influence

improvement of part yield from raw materials.

In summary, trade shows should be a focus in the promotion of scanning

technology. In addition, the importance of plant visits and peer discussions suggests that

manufacturers of scanning technology should also focus on installing a working system

in an industrial setting as word-of-mouth is an important factor. Finally, scanning

technology must be promoted to the secondary wood products industry as a tool for

increasing efficiency and influencing increased part yield in the plant as these were found

to be the most important affecting past scanning technology adoption in this industry. It is

thought that increased yield could be achieved with automated lumber grading

technology by providing an optimal lumber cut-up algorithm through more accurate

defect mapping and lumber grading.

103

REFERENCES

Adams, Larry. 1999. Cabinet Companies Step Out of the Kitchen. Wood and Wood Products. Vol. 104, No. 3, pages 51-63. Dempsey, Gilbert P. and W. G. Luppold. 1992. The State of Hardwood Lumber Markets. Northern Logger and Timber Processor. March, 3 pages. Hansen, Eric, V. S. Reddy, J. Punches, and R. Bush. 1995. Wood Material Use in the US Furniture Industry: 1993 and 1995. Center for Forest Products Marketing, Virginia Polytechnic Institute and State University. 19 pages. Hansen, Bruce and C. West. 1998. Trends in Domestic Hardwood Markets. Hardwood Symposium Proceedings. May 6-9, 9 pages. Mittal, Vikas, W. T. Ross, Jr., and P. M. Baldasare. 1998. The Asymmetric Impact of Negative and Positive Attribute-Level Performance on Overall Satisfaction and Repurchase Intentions. Journal of Marketing. Vol. 62, No. 1, pages 33-47. Punches, John W., E. N. Hansen, and R. J. Bush. 1995. Productivity Characteristics of the US Wood Cabinet Industry. Forest Products Journal. Vol. 45, No. 10, pages 33-38. Smith, Eric P. 1999. Professor of Statistics, Virginia Polytechnic Institute and State University, Statistical Consulting Center. Personal Interview, September 13, 1999. Taylor Russell E. 1999. South America (Wood Products Output/Exports to Soar). Presented at the World Wood Summit, Chicago, Illinois, May, 1999, 3 pages Vlosky, Richard P. 1996. Characteristics of US Hardwood Wood Component Manufacturers. Forest Products Journal. Vol. 46, No. 5, pages 37-43. Weinig, Michael. 1999. Parquet Flooring An Expanding Market. Wood and Wood Products. Vol. 104, No. 2, pages 89-90. West, Cynthia D. 1990. Competitive Determinants of Technology Diffusion in the Wood Household Furniture Industry. Virginia Polytechnic Institute and State University. April, pages 124-125.

104

CHAPTER 4

RESEARCH SUMMARY

Secondary manufacturers of wood products are currently facing problems such as

increased competition, and decreased availability of raw material. Problems such as

these are resulting in a trend toward increased automation and innovation in the

production process to increase plant efficiency and part yield, and to improve product

quality. The goal of this research was to assess, not quantify, the market for scanning

technology, specifically automated lumber grading technology, in the secondary wood

products industry. Scanning technology was defined in this research as a method of

optimizing the processing of wood through the use of various sensing components

coupled with computer memory. This research focused on automated lumber grading

technology. Automated lumber grading technology is defined in this research as a

system, which detects and recognizes wood defects and automatically assigns a grade to

lumber with little or no human assistance.

The research objectives were:

1. to identify potential adopters versus non-adopters of scanning technology,

2. to characterize the benefit bundle that would increase the marketable success of

automated lumber grading technology, and

3. to identify the best methods of promotion to this industry.

A mail survey was used to collect data from wood cabinet, furniture, dimension,

and flooring manufacturers. Approximately 1,200 questionnaires were mailed out to

these four sectors of the secondary wood products industry. A total of 439 questionnaires

105

were received, 209 of which were useable. Unusable questionnaires included

questionnaires for which addresses could not be located, companies no longer engaged in

activities relevant to this study, and questionnaires that were returned greater than 50%

incomplete. The resulting adjusted response rate was approximately 20%.

Demographic findings indicated that the average company in the sample frame

processed 7.3 million board feet of lumber per year, 98% reported annual sales of greater

than $1 million, 79% employed fewer than 200 people, 47% marketed their products

internationally, and 91% were privately owned. In addition, 70% of companies were not

using scanning technology at the time of this study, 76% utilized less than 6% of total

man-hours grading lumber, 82% required payback periods of less than 6 years, and the

mean cost to grade lumber per man-hour was $14.66. Dimension and flooring

manufacturers were found to process more lumber and spend both more time and money

in grading lumber when compared to smaller cabinet and furniture manufacturers.

Demographic information indicated that the cabinet and furniture samples were

comprised of mostly smaller companies in these industries.

The first objective of this research was to identify adopters versus non-adopters of

scanning technology. Dimension and flooring manufacturers perceived automated

lumber grading technology as more beneficial than smaller cabinet and furniture

manufacturers. This may be due to the variation in production processes among these

sectors. For example, the rough mill comprises the entire dimension and flooring

production process; therefore, the benefits of automated lumber grading technology may

be less difficult to see in these industries. Moreover, the relative small size of companies

106

in the cabinet and furniture samples may have influenced these findings. In addition, past

adopters of scanning technologies perceive tallying capability, sorting capability, and

production speed as more important product attributes than non-adopters when

considering automated lumber grading technology. These perceptions may indicate areas

where improvement or development in scanning technology is needed. Adopters are

more likely to be larger companies who market their products internationally. Due to the

high concentration of production activity in the rough mill and high lumber input

volumes, large dimension manufacturers may represent the best market for automated

lumber grading technology in the secondary wood products industry.

The second objective of this research was to characterize the benefit bundle that

will increase the overall marketable success of automated lumber grading technology in

the secondary wood products industry. The dimension and flooring industries showed an

overall greater interest in this technology compared to smaller cabinet and furniture

manufacturers. The dimension and flooring sectors rated machine attributes higher than

the cabinet and furniture sectors where significant differences were detected. This

research found overall important system attributes to the secondary wood products

industry including machine warranty, system accuracy, technical support, reduction in

labor costs, and machine durability. In addition, material sorting capability, species and

grade switching ability, compatibility with existing equipment, and color sorting

capability were of overall minor importance among the companies in the sample frame.

Technical support and equipment warranty from the manufacturer were desirable

attributes to all four sectors as well. The importance of service attributes such as these

107

indicate that manufacturers of scanning technologies must be attentive to customer needs

to be successful. In addition, the relative lack of sophisticated information systems

among many companies in the wood products industry may suggest that companies in

this industry expect service in the form of a human representative who provides on-site

service. Machine accuracy, ability to reduce labor costs, simplicity of operation, and

machine durability were included in the optimal benefit bundle for the majority of the

four sectors studied. Attributes included in the benefit bundles differed slightly among

the four sectors. The flooring sector was the only one to cite production speed and

tallying capability as attributes in the bundle. In addition, the furniture sector was the

only one to cite machine cost as an important attribute. Finally, the secondary wood

products industry expects automated lumber grading technology to cost no more than

$250,000. Moreover, automated lumber grading systems must be able to meet short

payback periods and be proven to increase part yield. Observing a system in use in an

industrial environment was cited by many as critical to their decision to invest in

automated lumber grading technology as well.

The third objective of this research was to identify the best methods of promotion

to the secondary wood products industry. Overall important methods of learning about

new technology among companies in the sample frame included, tradeshows, peer

discussions, and plant visits. While different mean rating values were reported, these

three methods were rated highest for each of the four industry sectors as well.

Advertisements, meetings and symposiums, scientific journals, and unsolicited sales

literature were all found to be of little importance in learning about new technology;

108

therefore, these methods would have little or no effect on promotion of scanning

technology to the secondary wood products industry. In addition, the dimension and

flooring industries demonstrated more open lines of communication between companies

compared to small cabinet and furniture manufacturers. This may indicate that word-of-

mouth promotion is important in this industry.

IMPLICATIONS TO THE SECONDARY WOOD PRODUCTS INDUSTRY

The secondary wood products industry is becoming increasingly high-tech in its

processing operations. Scanning technologies can already be found in rough-mill

optimized sawing systems and defect detection systems. Any place in the operation

where part yield can be increased and savings can be realized is being considered for

automation. This study shows that scanning technology, while in the development stages

in terms of automated lumber grading is becoming increasingly accepted in the industry.

However, analysis indicates that only those companies with sufficient capital and high

lumber grading costs will represent potential adopters of automated lumber grading

technology such as larger dimension manufacturers. Moreover, it is thought that

automated lumber grading technology may be better utilized in the primary wood

products industry; i.e., lumber could be machine graded at the sawmill and a bar code

attached containing the optimum cut-up algorithm. The bar code could be interpreted and

algorithm used in optimized cut-up operations in secondary processing eliminating the

need to grade lumber at the secondary processing facility.

Over half of the companies in the sample frame responded positively to the

benefits of automated lumber grading technology. While the majority of companies in

109

the sample frame did not have scanning technology at the time of this study, those who

did were larger companies. In addition, competition within the secondary wood products

industry has increased while becoming increasingly global. If one assumes that increased

utilization of scanning technology is in the future of wood processing automation and the

adoption of this technology can increase company competitiveness, it is logical to deduce

that smaller companies may be at a technological disadvantage in the future.

MARKETS FOR SCANNING TECHNOLOGY

From the perspective of manufacturers of scanning technology, the secondary

wood products industry appears to represent a potential market for automated lumber

grading technology. Specifically, large companies whose production process is

concentrated in the rough mill such as larger dimension manufacturers, represent

potential adopters of automated lumber grading technology. In addition, the primary

wood processing industry may represent a potential market for automated lumber grading

technology. It is thought that primary processors could offer their secondary wood

processing customers a value-added product in machine-graded lumber that included an

optimal bar-coded cut-up algorithm.

The current trend toward increased plant automation provides potential for

increased innovation of all types in the industry. In addition, the important role of word-

of-mouth promotion and the open lines of communication found to exist in the industry

indicate that there is potential for success for a company who can deliver an effective

system operating in an industrial environment.

110

AREAS OF FUTURE RESEARCH

This research has shown that a market for automated lumber grading technology

does exist in the secondary wood products industry. This research indicates that the

majority of this industry has not adopted scanning technology, and that a clear distinction

exists between industry sectors on the perception that automated lumber grading

technology would be beneficial to their industry. Moreover, adopters of this technology

will be characterized by large company size and relatively high lumber grading costs. In

addition, this research shows that slight differences exist between industry sectors in the

machine capabilities that are important to them, and the most effective methods of

promoting this technology to these sectors of the industry. It should be noted, however,

that conclusions are skewed toward smaller cabinet and furniture manufacturers.

This research addressed fourteen system attributes and nine promotion methods.

Other factors could be applied to broaden the scope. Machine attributes could be

segmented into factors related to the machine, factors related to the manufacturer, and

factors related to the wood raw material. Factorial analysis could be used to delineate

these and other categories. In addition, research could focus more on the decision-

makers in each industry sector, in terms of how they receive information and how to relay

information to them. Moreover, it is thought that the number of professional certified

lumber graders is diminishing. As certified lumber graders become more difficult to find,

the cost of not grading lumber becomes an issue. No research has addressed this issue;

therefore, future studies could focus on the cost of not grading lumber in the secondary

wood products industry.

111

LIMITATIONS TO THE STUDY

As with most research, not every aspect can be identified or anticipated. This

study was developed to assess the market for scanning technology in the secondary wood

products industry, using automated lumber grading technology as a case study. The fact

that the perception of scanning technology in the US is influenced by scanning

manufacturers marketing systems in different regions could influence the results of this

study. Personal interviews with persons in the industry indicate that perception of

scanning technology in the secondary wood products industry is often influenced by the

perception of the manufacturer with which one is familiar rather than the actual

capabilities of the technology.

In addition, the sample response rate was lower than anticipated, particularly for

the cabinets and furniture samples. This could have been influenced by the frequency

with which these sectors of the industry are surveyed. In addition, it appears from our

demographic information that the cabinet and furniture samples were comprised of

predominantly smaller producers. This makes our conclusions more applicable to the

smaller producers in these sectors of the industry.

112

APPENDIX A: SURVEY QUESTIONNAIRE

Scanning Technology in the Secondary Wood Products Industry The forest products industry is becoming a high-tech industry. As technology advances, many wood products firms are finding ways to automate their operations, increase efficiency and yield, and improve quality. However, to adequately provide the technology that is beneficial to wood products firms, manufacturers of high technology equipment must fully understand their needs. This questionnaire is designed to better understand the factors that are important to users of this technology. This information will allow manufacturers to produce scanning technologies that better meet your needs. For the purpose of this research, scanning technology is defined as: camera, laser, x-ray, etc. used in cut-up operations, defect detection, quality control scanning, or color matching. 1. Which category best describes the nature of your

business? (Please check the appropriate box.)

Cabinet manufacturer Furniture manufacturer Dimension mill Flooring manufacturer

2. Do you currently have scanning technology in

place?

No If No, please go to question 4. Yes (Please describe.)

3. If you currently have scanning technology in place,

what factors influenced the decision to invest in the equipment? (Please check all that apply.)

Increased efficiency Automation Yield Accuracy Other (Please list.)____________________

Lumber Grading 4. When in your operation is lumber graded? (Please

check all that apply.)

At delivery First-In-First-Out Before drying if applicable After drying if applicable

5. Please rank the following factors on their

importance to lumber grading in your company. (1=least important, 7=most important)

Least Average Most Important Importance Important Invoice verification 1 2 3 4 5 6 7 Better matching of grade to product 1 2 3 4 5 6 7 Facilitates lumber sorting 1 2 3 4 5 6 7 Other (Please list.)______ 1 2 3 4 5 6 7 6. What would you estimate your average cost per

man-hour of grading lumber to be? $__________________/hour 7. How many man-hours would you estimate grading

lumber requires in your company as a percentage of total man-hours?

Less than 2% 3% - 5% 6% - 9% Greater than 10%

113

A general definition of scanning technology was given in the introduction (camera, laser, x-ray scanning, etc. used in cut-up operations, defect detection, quality control, etc). The next section introduces automated lumber grading systems defined as: systems using camera scanning mechanisms and programmable computer software to assign a grade to lumber. 8. How important would the following factors be in

your decision to purchase an automated lumber grading system? (1=least important, 7=most important)

Least Average Most Important Importance Important Accuracy 1 2 3 4 5 6 7

Cost 1 2 3 4 5 6 7

Durability 1 2 3 4 5 6 7

Compatibility with existing equipment 1 2 3 4 5 6 7

Speed 1 2 3 4 5 6 7

Ability to quickly switch grading rules 1 2 3 4 5 6 7

Ability to quickly switch species 1 2 3 4 5 6 7

Simplicity of operation 1 2 3 4 5 6 7

Technical support 1 2 3 4 5 6 7

Equipment warranty 1 2 3 4 5 6 7

Reduction in labor costs 1 2 3 4 5 6 7

Tallying capabilities 1 2 3 4 5 6 7

Sorting capabilities 1 2 3 4 5 6 7

Color sorting capabilities 1 2 3 4 5 6 7 9. Do you believe that an automated lumber grading

system would benefit your business?

No Yes

10. How much would you expect an automated lumber grading system with all desired features to cost? (Please check one.)

Less than $50,000 $50,000 - $150,000 $150,001 - $250,000 $250,001 - $500,000 Greater than $500,000

11. What would be your expected payback period for

an automated lumber grading system? (Please check one.)

Less than 2 years 3 –5 years 6 – 10 years Greater than 10 years

12. Has a manufacturer of automated grading systems

(AGS) ever approached you concerning purchasing an AGS?

No Yes

13. Do companies within your industry ever contact

you to inquire about high technology equipment you have or have not installed?

No Yes

14. How important to your company are the following

in learning about new technology and equipment? (1=least important, 7=most important)

Least Average Most Important Importance Important Advertisements 1 2 3 4 5 6 7 Trade shows 1 2 3 4 5 6 7 Plant visits 1 2 3 4 5 6 7 Personal sales calls from manufacturers 1 2 3 4 5 6 7 Meetings, Symposiums 1 2 3 4 5 6 7 Trade journals 1 2 3 4 5 6 7 Scientific journals 1 2 3 4 5 6 7 Peer discussions 1 2 3 4 5 6 7 Unsolicited sales 1 2 3 4 5 6 7 literature Other: __________ 1 2 3 4 5 6 7 (Please specify.)

114

15. Does your company wish to be the industry leader in adopting new technology?

No Yes

16. Do you have a professional engineer on staff?

No Yes

Company Demographics 17. What would you estimate your company’s annual lumber input volume to be in board feet? _______________board feet 18. Do you market your products internationally?

No Yes

19. What were the total gross sales for your company

in 1998? (Please check one.)

Less than $100,000 $100,000 - $500,000 $500,001 - $1,000,000 $1,000,001 - $5,000,000 $5,000,001 - $10,000,000 Greater than $10,000,000

20. Is your company publicly or privately owned?

Public Private

21. How many full-time employees work for your

company? (Please check one.)

Fewer than 25 25 – 50 51 – 100 101 – 200 201 – 300 301 – 400 Greater than 400

22. Given that the technology is proven effective and available, what would prevent you from purchasing an automated lumber grading system?

23. Ultimately, what would convince you to purchase

an automated lumber grading system? 24. Please cite specific reasons why you are, or are not

satisfied with technology equipment currently in place, if applicable.

25. Is there anything else you wish to share with us

concerning scanning technology, or lumber grading?

Thank you for your help! Please fold the questionnaire, such that the return address is visible, and return by mail. The postage is prepaid.

115

APPENDIX B: SURVEY COVER LETTER

116

APPENDIX C: REMINDER POSTCARD

117

VITAE

Dan W. Cumbo II, son of Danny W. Cumbo Sr. and Ruth S. Hankins, was born in Danville, Virginia on

October 19, 1971. After graduating from George Washington High School in 1989, Mr. Cumbo attended Danville

Community College where he received a diploma in Graphic Arts and Printing Technology in 1991. Mr. Cumbo

returned to Danville Community College in 1993 to the Science curriculum. In 1995 Mr. Cumbo transferred to

Virginia Polytechnic Institute and State University where he received a Bachelor of Science degree in Wood Science

& Forest Products in 1998. Mr. Cumbo continued his education by earning a Master of Science degree in Wood

Science & Forest Products from Virginia Polytechnic Institute and State University in 1999. Mr. Cumbo will begin

his career in the wood products industry as Project Manager with Visador Company in Marion, Virginia.